Allow failover for replicas

Revert "Do not unban replicas if a primary is available (#843)"

This reverts commit cdcfa99fb9.

.circleci/config.yml

@@ -63,6 +63,9 @@ jobs:
       - run:
           name: "Lint"
           command: "cargo fmt --check"
+      - run:
+          name: "Clippy"
+          command: "cargo clippy --all --all-targets -- -Dwarnings"
       - run:
           name: "Tests"
           command: "cargo clean && cargo build && cargo test && bash .circleci/run_tests.sh && .circleci/generate_coverage.sh"

.circleci/pgcat.toml

@@ -59,6 +59,7 @@ admin_password = "admin_pass"
 # session: one server connection per connected client
 # transaction: one server connection per client transaction
 pool_mode = "transaction"
+prepared_statements_cache_size = 500
 
 # If the client doesn't specify, route traffic to
 # this role by default.
@@ -141,6 +142,7 @@ query_parser_enabled = true
 query_parser_read_write_splitting = true
 primary_reads_enabled = true
 sharding_function = "pg_bigint_hash"
+prepared_statements_cache_size = 500
 
 [pools.simple_db.users.0]
 username = "simple_user"

.circleci/run_tests.sh

@@ -26,6 +26,7 @@ PGPASSWORD=sharding_user pgbench -h 127.0.0.1 -U sharding_user shard1 -i
 PGPASSWORD=sharding_user pgbench -h 127.0.0.1 -U sharding_user shard2 -i
 
 # Start Toxiproxy
+kill -9 $(pgrep toxiproxy) || true
 LOG_LEVEL=error toxiproxy-server &
 sleep 1
 
@@ -106,10 +107,26 @@ cd ../..
 # These tests will start and stop the pgcat server so it will need to be restarted after the tests
 #
 pip3 install -r tests/python/requirements.txt
-python3 tests/python/tests.py || exit 1
+pytest || exit 1
+
+
+#
+# Go tests
+# Starts its own pgcat server
+#
+pushd tests/go
+/usr/local/go/bin/go test || exit 1
+popd
 
 start_pgcat "info"
 
+#
+# Rust tests
+#
+cd tests/rust
+cargo run
+cd ../../
+
 # Admin tests
 export PGPASSWORD=admin_pass
 psql -U admin_user -e -h 127.0.0.1 -p 6432 -d pgbouncer -c 'SHOW STATS' > /dev/null
@@ -161,3 +178,6 @@ killall pgcat -s SIGINT
 
 # Allow for graceful shutdown
 sleep 1
+
+kill -9 $(pgrep toxiproxy)
+sleep 1

.github/dependabot.yml

@@ -10,3 +10,7 @@ updates:
     commit-message:
       prefix: "chore(deps)"
     open-pull-requests-limit: 10
+  - package-ecosystem: "github-actions"
+    directory: "/"
+    schedule:
+      interval: "weekly"

.github/workflows/build-and-push.yaml

@@ -2,7 +2,9 @@ name: Build and Push
 
 on:
   push:
-    branches:    
+    paths:
+      - '!charts/**.md'
+    branches:
       - main
     tags:
       - v*
@@ -21,14 +23,17 @@ jobs:
 
     steps:
       - name: Checkout Repository
-        uses: actions/checkout@v3
+        uses: actions/checkout@v4
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
 
       - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v2
+        uses: docker/setup-buildx-action@v3
 
       - name: Determine tags
         id: metadata
-        uses: docker/metadata-action@v4
+        uses: docker/metadata-action@v5
         with:
           images: ${{ env.registry }}/${{ env.image-name }}
           tags: |
@@ -40,15 +45,18 @@ jobs:
             type=raw,value=latest,enable={{ is_default_branch }}
 
       - name: Log in to the Container registry
-        uses: docker/login-action@v2.1.0
+        uses: docker/login-action@v3
         with:
           registry: ${{ env.registry }}
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Build and push ${{ env.image-name }}
-        uses: docker/build-push-action@v3
+        uses: docker/build-push-action@v6
         with:
+          context: .
+          platforms: linux/amd64,linux/arm64
+          provenance: false
           push: true
           tags: ${{ steps.metadata.outputs.tags }}
           labels: ${{ steps.metadata.outputs.labels }}

.github/workflows/chart-lint-test.yaml

@@ -0,0 +1,50 @@
+name: Lint and Test Charts
+
+on:
+  pull_request:
+    paths:
+      - charts/**
+      - '!charts/**.md'
+jobs:
+  lint-test:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v3.1.0
+        with:
+          fetch-depth: 0
+
+      - name: Set up Helm
+        uses: azure/setup-helm@v3
+        with:
+          version: v3.8.1
+
+      # Python is required because `ct lint` runs Yamale (https://github.com/23andMe/Yamale) and
+      # yamllint (https://github.com/adrienverge/yamllint) which require Python
+      - name: Set up Python
+        uses: actions/setup-python@v5.1.0
+        with:
+          python-version: 3.7
+
+      - name: Set up chart-testing
+        uses: helm/chart-testing-action@v2.2.1
+        with:
+          version: v3.5.1
+
+      - name: Run chart-testing (list-changed)
+        id: list-changed
+        run: |
+          changed=$(ct list-changed --config ct.yaml)
+          if [[ -n "$changed" ]]; then
+            echo "changed=true" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Run chart-testing (lint)
+        run: ct lint --config ct.yaml
+
+      - name: Create kind cluster
+        uses: helm/kind-action@v1.10.0
+        if: steps.list-changed.outputs.changed == 'true'
+
+      - name: Run chart-testing (install)
+        run: ct install --config ct.yaml

.github/workflows/chart-release.yaml

@@ -0,0 +1,40 @@
+name: Release Charts
+
+on:
+  push:
+    paths:
+      - charts/**
+      - '!**.md'
+    branches:
+      - main
+
+jobs:
+  release:
+    runs-on: ubuntu-latest
+
+    permissions:
+      contents: write
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@8ade135a41bc03ea155e62e844d188df1ea18608 # v4.1.0
+        with:
+          fetch-depth: 0
+
+      - name: Configure Git
+        run: |
+          git config user.name "$GITHUB_ACTOR"
+          git config user.email "$GITHUB_ACTOR@users.noreply.github.com"
+
+      - name: Install Helm
+        uses: azure/setup-helm@5119fcb9089d432beecbf79bb2c7915207344b78 # v3.5
+        with:
+          version: v3.13.0
+
+      - name: Run chart-releaser
+        uses: helm/chart-releaser-action@a917fd15b20e8b64b94d9158ad54cd6345335584 # v1.6.0
+        with:
+          charts_dir: charts
+          config: cr.yaml
+        env:
+          CR_TOKEN: "${{ secrets.GITHUB_TOKEN }}"

.github/workflows/generate-chart-readme.yaml

@@ -0,0 +1,48 @@
+name: '[CI/CD] Update README metadata'
+
+on:
+  pull_request_target:
+    branches:
+      - main
+    paths:
+      - 'charts/*/values.yaml'
+# Remove all permissions by default
+permissions: {}
+jobs:
+  update-readme-metadata:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - name: Install readme-generator-for-helm
+        run: npm install -g @bitnami/readme-generator-for-helm
+      - name: Checkout
+        uses: actions/checkout@8ade135a41bc03ea155e62e844d188df1ea18608
+        with:
+          path: charts
+          ref: ${{github.event.pull_request.head.ref}}
+          repository: ${{github.event.pull_request.head.repo.full_name}}
+          token: ${{ secrets.GITHUB_TOKEN }}
+      - name: Execute readme-generator-for-helm
+        env:
+          DIFF_URL: "${{github.event.pull_request.diff_url}}"
+          TEMP_FILE: "${{runner.temp}}/pr-${{github.event.number}}.diff"
+        run: |
+          # This request doesn't consume API calls.
+          curl -Lkso $TEMP_FILE $DIFF_URL
+          files_changed="$(sed -nr 's/[\-\+]{3} [ab]\/(.*)/\1/p' $TEMP_FILE | sort | uniq)"
+          # Adding || true to avoid "Process exited with code 1" errors
+          charts_dirs_changed="$(echo "$files_changed" | xargs dirname | grep -o "pgcat/[^/]*" | sort | uniq || true)"
+          for chart in ${charts_dirs_changed}; do
+            echo "Updating README.md for ${chart}"
+            readme-generator --values "charts/${chart}/values.yaml" --readme "charts/${chart}/README.md" --schema "/tmp/schema.json"
+          done
+      - name: Push changes
+        run: |
+          # Push all the changes
+          cd charts
+          if git status -s | grep pgcat; then
+            git config user.name "$GITHUB_ACTOR"
+            git config user.email "$GITHUB_ACTOR@users.noreply.github.com"
+            git add . && git commit -am "Update README.md with readme-generator-for-helm" --signoff && git push
+          fi

.github/workflows/publish-deb-package.yml

@@ -1,10 +1,13 @@
 name: pgcat package (deb)
 
 on:
+  push:
+    tags:
+      - v*
   workflow_dispatch:
     inputs:
       packageVersion:
-        default: "1.1.2-dev"
+        default: "1.1.2-dev1"
 jobs:
   build:
     strategy:
@@ -16,6 +19,14 @@ jobs:
     runs-on: ${{ matrix.os }}
     steps:
     - uses: actions/checkout@v3
+    - name: Set package version
+      if: github.event_name == 'push'  # For push event
+      run: |
+        TAG=${{ github.ref_name }}
+        echo "packageVersion=${TAG#v}" >> "$GITHUB_ENV"
+    - name: Set package version (manual dispatch)
+      if: github.event_name == 'workflow_dispatch'  # For manual dispatch
+      run: echo "packageVersion=${{ github.event.inputs.packageVersion }}" >> "$GITHUB_ENV"
     - uses: actions-rs/toolchain@v1
       with:
         toolchain: stable
@@ -39,10 +50,10 @@ jobs:
           export ARCH=arm64
         fi
 
-        bash utilities/deb.sh ${{ inputs.packageVersion }}
+        bash utilities/deb.sh ${{ env.packageVersion }}
 
         deb-s3 upload \
           --lock \
           --bucket apt.postgresml.org \
-          pgcat-${{ inputs.packageVersion }}-ubuntu22.04-${ARCH}.deb \
+          pgcat-${{ env.packageVersion }}-ubuntu22.04-${ARCH}.deb \
           --codename $(lsb_release -cs)

.gitignore

@@ -10,3 +10,6 @@ lcov.info
 dev/.bash_history
 dev/cache
 !dev/cache/.keepme
+.venv
+**/__pycache__
+.bundle

CONFIG.md

@@ -36,10 +36,11 @@ Port at which prometheus exporter listens on.
 ### connect_timeout
 ```
 path: general.connect_timeout
-default: 5000 # milliseconds
+default: 1000 # milliseconds
 ```
 
-How long to wait before aborting a server connection (ms).
+How long the client waits to obtain a server connection before aborting (ms).
+This is similar to PgBouncer's `query_wait_timeout`.
 
 ### idle_timeout
 ```
@@ -259,22 +260,6 @@ Password to be used for connecting to servers to obtain the hash used for md5 au
 specified in `auth_query_user`. The connection will be established using the database configured in the pool.
 This parameter is inherited by every pool and can be redefined in pool configuration.
 
-### prepared_statements
-```
-path: general.prepared_statements
-default: false
-```
-
-Whether to use prepared statements or not.
-
-### prepared_statements_cache_size
-```
-path: general.prepared_statements_cache_size
-default: 500
-```
-
-Size of the prepared statements cache.
-
 ### dns_cache_enabled
 ```
 path: general.dns_cache_enabled
@@ -324,6 +309,24 @@ If the client doesn't specify, PgCat routes traffic to this role by default.
 `replica` round-robin between replicas only without touching the primary,
 `primary` all queries go to the primary unless otherwise specified.
 
+### replica_to_primary_failover_enabled
+```
+path: pools.<pool_name>.replica_to_primary_failover_enabled
+default: "false"
+```
+
+If set to true, when the specified role is `replica` (either by setting `default_role` or manually)
+and all replicas are banned, queries will be sent to the primary (until a replica is back online).
+
+### prepared_statements_cache_size
+```
+path: general.prepared_statements_cache_size
+default: 0
+```
+
+Size of the prepared statements cache. 0 means disabled.
+TODO: update documentation
+
 ### query_parser_enabled
 ```
 path: pools.<pool_name>.query_parser_enabled
@@ -469,10 +472,18 @@ path: pools.<pool_name>.users.<user_index>.pool_size
 default: 9
 ```
 
-Maximum number of server connections that can be established for this user
+Maximum number of server connections that can be established for this user.
 The maximum number of connection from a single Pgcat process to any database in the cluster
 is the sum of pool_size across all users.
 
+### min_pool_size
+```
+path: pools.<pool_name>.users.<user_index>.min_pool_size
+default: 0
+```
+
+Minimum number of idle server connections to retain for this pool.
+
 ### statement_timeout
 ```
 path: pools.<pool_name>.users.<user_index>.statement_timeout
@@ -482,6 +493,16 @@ default: 0
 Maximum query duration. Dangerous, but protects against DBs that died in a non-obvious way.
 0 means it is disabled.
 
+### connect_timeout
+```
+path: pools.<pool_name>.users.<user_index>.connect_timeout
+default: <UNSET> # milliseconds
+```
+
+How long the client waits to obtain a server connection before aborting (ms).
+This is similar to PgBouncer's `query_wait_timeout`.
+If unset, uses the `connect_timeout` defined globally.
+
 ## `pools.<pool_name>.shards.<shard_index>` Section
 
 ### servers
@@ -509,4 +530,3 @@ default: "shard0"
 ```
 
 Database name (e.g. "postgres")
-

CONTRIBUTING.md

@@ -2,10 +2,36 @@
 
 Thank you for contributing! Just a few tips here:
 
-1. `cargo fmt` your code before opening up a PR
+1. `cargo fmt` and `cargo clippy` your code before opening up a PR
 2. Run the test suite (e.g. `pgbench`) to make sure everything still works. The tests are in `.circleci/run_tests.sh`.
 3. Performance is important, make sure there are no regressions in your branch vs. `main`.
 
+## How to run the integration tests locally and iterate on them
+We have integration tests written in Ruby, Python, Go and Rust.
+Below are the steps to run them in a developer-friendly way that allows iterating and quick turnaround.
+Hear me out, this should be easy, it will involve opening a shell into a container with all the necessary dependancies available for you and you can modify the test code and immediately rerun your test in the interactive shell.
+
+
+Quite simply, make sure you have docker installed and then run
+`./start_test_env.sh`
+
+That is it!
+
+Within this test environment you can modify the file in your favorite IDE and rerun the tests without having to bootstrap the entire environment again.
+
+Once the environment is ready, you can run the tests by running
+Ruby:   `cd /app/tests/ruby && bundle exec ruby <test_name>.rb --format documentation`
+Python: `cd /app/ && pytest`
+Rust:   `cd /app/tests/rust && cargo run`
+Go:     `cd /app/tests/go && /usr/local/go/bin/go test`
+
+You can also rebuild PgCat directly within the environment and the tests will run against the newly built binary
+To rebuild PgCat, just run `cargo build` within the container under `/app`
+
+![Animated gif showing how to run tests](https://github.com/user-attachments/assets/2258fde3-2aed-4efb-bdc5-e4f12dcd4d33)
+
+
+
 Happy hacking!
 
 ## TODOs

Cargo.lock

@@ -17,6 +17,17 @@ version = "1.0.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f26201604c87b1e01bd3d98f8d5d9a8fcbb815e8cedb41ffccbeb4bf593a35fe"
 
+[[package]]
+name = "ahash"
+version = "0.8.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2c99f64d1e06488f620f932677e24bc6e2897582980441ae90a671415bd7ec2f"
+dependencies = [
+ "cfg-if",
+ "once_cell",
+ "version_check",
+]
+
 [[package]]
 name = "aho-corasick"
 version = "1.0.2"
@@ -26,6 +37,12 @@ dependencies = [
  "memchr",
 ]
 
+[[package]]
+name = "allocator-api2"
+version = "0.2.16"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0942ffc6dcaadf03badf6e6a2d0228460359d5e34b57ccdc720b7382dfbd5ec5"
+
 [[package]]
 name = "android-tzdata"
 version = "0.1.1"
@@ -129,6 +146,12 @@ dependencies = [
  "syn 2.0.26",
 ]
 
+[[package]]
+name = "atomic-waker"
+version = "1.1.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1505bd5d3d116872e7271a6d4e16d81d0c8570876c8de68093a09ac269d8aac0"
+
 [[package]]
 name = "atomic_enum"
 version = "0.2.0"
@@ -169,12 +192,11 @@ checksum = "604178f6c5c21f02dc555784810edfb88d34ac2c73b2eae109655649ee73ce3d"
 
 [[package]]
 name = "bb8"
-version = "0.8.1"
+version = "0.8.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "98b4b0f25f18bcdc3ac72bdb486ed0acf7e185221fd4dc985bc15db5800b0ba2"
+checksum = "d89aabfae550a5c44b43ab941844ffcd2e993cb6900b342debf59e9ea74acdb8"
 dependencies = [
  "async-trait",
- "futures-channel",
  "futures-util",
  "parking_lot",
  "tokio",
@@ -525,34 +547,32 @@ checksum = "b6c80984affa11d98d1b88b66ac8853f143217b399d3c74116778ff8fdb4ed2e"
 
 [[package]]
 name = "h2"
-version = "0.3.20"
+version = "0.4.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "97ec8491ebaf99c8eaa73058b045fe58073cd6be7f596ac993ced0b0a0c01049"
+checksum = "524e8ac6999421f49a846c2d4411f337e53497d8ec55d67753beffa43c5d9205"
 dependencies = [
+ "atomic-waker",
  "bytes",
  "fnv",
  "futures-core",
  "futures-sink",
- "futures-util",
  "http",
- "indexmap 1.9.3",
+ "indexmap",
  "slab",
  "tokio",
  "tokio-util",
  "tracing",
 ]
 
-[[package]]
-name = "hashbrown"
-version = "0.12.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8a9ee70c43aaf417c914396645a0fa852624801b24ebb7ae78fe8272889ac888"
-
 [[package]]
 name = "hashbrown"
 version = "0.14.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "2c6201b9ff9fd90a5a3bac2e56a830d0caa509576f0e503818ee82c181b3437a"
+dependencies = [
+ "ahash",
+ "allocator-api2",
+]
 
 [[package]]
 name = "heck"
@@ -588,9 +608,9 @@ dependencies = [
 
 [[package]]
 name = "http"
-version = "0.2.9"
+version = "1.1.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bd6effc99afb63425aff9b05836f029929e345a6148a14b7ecd5ab67af944482"
+checksum = "21b9ddb458710bc376481b842f5da65cdf31522de232c1ca8146abce2a358258"
 dependencies = [
  "bytes",
  "fnv",
@@ -599,12 +619,24 @@ dependencies = [
 
 [[package]]
 name = "http-body"
-version = "0.4.5"
+version = "1.0.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d5f38f16d184e36f2408a55281cd658ecbd3ca05cce6d6510a176eca393e26d1"
+checksum = "1efedce1fb8e6913f23e0c92de8e62cd5b772a67e7b3946df930a62566c93184"
 dependencies = [
  "bytes",
  "http",
+]
+
+[[package]]
+name = "http-body-util"
+version = "0.1.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "793429d76616a256bcb62c2a2ec2bed781c8307e797e2598c50010f2bee2544f"
+dependencies = [
+ "bytes",
+ "futures-util",
+ "http",
+ "http-body",
  "pin-project-lite",
 ]
 
@@ -622,13 +654,12 @@ checksum = "c4a1e36c821dbe04574f602848a19f742f4fb3c98d40449f11bcad18d6b17421"
 
 [[package]]
 name = "hyper"
-version = "0.14.27"
+version = "1.4.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ffb1cfd654a8219eaef89881fdb3bb3b1cdc5fa75ded05d6933b2b382e395468"
+checksum = "50dfd22e0e76d0f662d429a5f80fcaf3855009297eab6a0a9f8543834744ba05"
 dependencies = [
  "bytes",
  "futures-channel",
- "futures-core",
  "futures-util",
  "h2",
  "http",
@@ -637,13 +668,26 @@ dependencies = [
  "httpdate",
  "itoa",
  "pin-project-lite",
- "socket2 0.4.9",
+ "smallvec",
  "tokio",
- "tower-service",
- "tracing",
  "want",
 ]
 
+[[package]]
+name = "hyper-util"
+version = "0.1.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "cde7055719c54e36e95e8719f95883f22072a48ede39db7fc17a4e1d5281e9b9"
+dependencies = [
+ "bytes",
+ "futures-util",
+ "http",
+ "http-body",
+ "hyper",
+ "pin-project-lite",
+ "tokio",
+]
+
 [[package]]
 name = "iana-time-zone"
 version = "0.1.57"
@@ -688,16 +732,6 @@ dependencies = [
  "unicode-normalization",
 ]
 
-[[package]]
-name = "indexmap"
-version = "1.9.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bd070e393353796e801d209ad339e89596eb4c8d430d18ede6a1cced8fafbd99"
-dependencies = [
- "autocfg",
- "hashbrown 0.12.3",
-]
-
 [[package]]
 name = "indexmap"
 version = "2.0.0"
@@ -705,7 +739,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "d5477fe2230a79769d8dc68e0eabf5437907c0457a5614a9e8dddb67f65eb65d"
 dependencies = [
  "equivalent",
- "hashbrown 0.14.0",
+ "hashbrown",
 ]
 
 [[package]]
@@ -821,6 +855,15 @@ version = "0.4.19"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b06a4cde4c0f271a446782e3eff8de789548ce57dbc8eca9292c27f4a42004b4"
 
+[[package]]
+name = "lru"
+version = "0.12.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "1efa59af2ddfad1854ae27d75009d538d0998b4b2fd47083e743ac1a10e46c60"
+dependencies = [
+ "hashbrown",
+]
+
 [[package]]
 name = "lru-cache"
 version = "0.1.2"
@@ -990,7 +1033,7 @@ checksum = "9b2a4787296e9989611394c33f193f676704af1686e70b8f8033ab5ba9a35a94"
 
 [[package]]
 name = "pgcat"
-version = "1.1.2-dev"
+version = "1.2.0"
 dependencies = [
  "arc-swap",
  "async-trait",
@@ -1004,10 +1047,13 @@ dependencies = [
  "fallible-iterator",
  "futures",
  "hmac",
+ "http-body-util",
  "hyper",
+ "hyper-util",
  "itertools",
  "jemallocator",
  "log",
+ "lru",
  "md-5",
  "nix",
  "num_cpus",
@@ -1447,9 +1493,9 @@ dependencies = [
 
 [[package]]
 name = "smallvec"
-version = "1.11.0"
+version = "1.13.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "62bb4feee49fdd9f707ef802e22365a35de4b7b299de4763d44bfea899442ff9"
+checksum = "3c5e1a9a646d36c3599cd173a41282daf47c44583ad367b8e6837255952e5c67"
 
 [[package]]
 name = "socket2"
@@ -1479,9 +1525,9 @@ checksum = "6e63cff320ae2c57904679ba7cb63280a3dc4613885beafb148ee7bf9aa9042d"
 
 [[package]]
 name = "sqlparser"
-version = "0.34.0"
+version = "0.41.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "37d3706eefb17039056234df6b566b0014f303f867f2656108334a55b8096f59"
+checksum = "5cc2c25a6c66789625ef164b4c7d2e548d627902280c13710d33da8222169964"
 dependencies = [
  "log",
  "sqlparser_derive",
@@ -1489,13 +1535,13 @@ dependencies = [
 
 [[package]]
 name = "sqlparser_derive"
-version = "0.1.1"
+version = "0.2.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "55fe75cb4a364c7f7ae06c7dbbc8d84bddd85d6cdf9975963c3935bc1991761e"
+checksum = "01b2e185515564f15375f593fb966b5718bc624ba77fe49fa4616ad619690554"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 1.0.109",
+ "syn 2.0.26",
 ]
 
 [[package]]
@@ -1710,19 +1756,13 @@ version = "0.19.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "f8123f27e969974a3dfba720fdb560be359f57b44302d280ba72e76a74480e8a"
 dependencies = [
- "indexmap 2.0.0",
+ "indexmap",
  "serde",
  "serde_spanned",
  "toml_datetime",
  "winnow",
 ]
 
-[[package]]
-name = "tower-service"
-version = "0.3.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b6bc1c9ce2b5135ac7f93c72918fc37feb872bdc6a5533a8b85eb4b86bfdae52"
-
 [[package]]
 name = "tracing"
 version = "0.1.37"

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "pgcat"
-version = "1.1.2-dev"
+version = "1.2.0"
 edition = "2021"
 
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
@@ -8,7 +8,7 @@ edition = "2021"
 tokio = { version = "1", features = ["full"] }
 bytes = "1"
 md-5 = "0.10"
-bb8 = "0.8.1"
+bb8 = "=0.8.6"
 async-trait = "0.1"
 rand = "0.8"
 chrono = "0.4"
@@ -19,7 +19,7 @@ serde_derive = "1"
 regex = "1"
 num_cpus = "1"
 once_cell = "1"
-sqlparser = {version = "0.34", features = ["visitor"] }
+sqlparser = { version = "0.41", features = ["visitor"] }
 log = "0.4"
 arc-swap = "1"
 parking_lot = "0.12.1"
@@ -29,7 +29,9 @@ base64 = "0.21"
 stringprep = "0.1"
 tokio-rustls = "0.24"
 rustls-pemfile = "1"
-hyper = { version = "0.14", features = ["full"] }
+http-body-util = "0.1.2"
+hyper = { version = "1.4.1", features = ["full"] }
+hyper-util = { version = "0.1.7", features = ["tokio"] }
 phf = { version = "0.11.1", features = ["macros"] }
 exitcode = "1.1.2"
 futures = "0.3"
@@ -47,8 +49,12 @@ serde_json = "1"
 itertools = "0.10"
 clap = { version = "4.3.1", features = ["derive", "env"] }
 tracing = "0.1.37"
-tracing-subscriber = { version = "0.3.17", features = ["json", "env-filter", "std"]}
+tracing-subscriber = { version = "0.3.17", features = [
+    "json",
+    "env-filter",
+    "std",
+] }
+lru = "0.12.0"
 
 [target.'cfg(not(target_env = "msvc"))'.dependencies]
 jemallocator = "0.5.0"
-

Dockerfile

@@ -1,4 +1,4 @@
-FROM rust:1-slim-bookworm AS builder
+FROM rust:1.79.0-slim-bookworm AS builder
 
 RUN apt-get update && \
     apt-get install -y build-essential
@@ -8,8 +8,15 @@ WORKDIR /app
 RUN cargo build --release
 
 FROM debian:bookworm-slim
+RUN apt-get update && apt-get install  -o Dpkg::Options::=--force-confdef -yq --no-install-recommends \
+    postgresql-client \
+    # Clean up layer
+    && apt-get clean \
+    && rm -rf /var/lib/apt/lists/* /tmp/* /var/tmp/* \
+    && truncate -s 0 /var/log/*log
 COPY --from=builder /app/target/release/pgcat /usr/bin/pgcat
 COPY --from=builder /app/pgcat.toml /etc/pgcat/pgcat.toml
 WORKDIR /etc/pgcat
 ENV RUST_LOG=info
 CMD ["pgcat"]
+STOPSIGNAL SIGINT

Dockerfile.ci

@@ -1,4 +1,4 @@
-FROM cimg/rust:1.67.1
+FROM cimg/rust:1.79.0
 COPY --from=sclevine/yj /bin/yj /bin/yj
 RUN /bin/yj -h
 RUN sudo apt-get update && \
@@ -9,6 +9,9 @@ RUN sudo apt-get update && \
 	sudo apt-get upgrade curl && \
 	cargo install cargo-binutils rustfilt && \
 	rustup component add llvm-tools-preview && \
- 	pip3 install psycopg2 && sudo gem install bundler && \
+	pip3 install psycopg2 && sudo gem install bundler && \
 	wget -O /tmp/toxiproxy-2.4.0.deb https://github.com/Shopify/toxiproxy/releases/download/v2.4.0/toxiproxy_2.4.0_linux_$(dpkg --print-architecture).deb && \
 	sudo dpkg -i /tmp/toxiproxy-2.4.0.deb
+RUN wget -O /tmp/go1.21.3.linux-$(dpkg --print-architecture).tar.gz https://go.dev/dl/go1.21.3.linux-$(dpkg --print-architecture).tar.gz && \
+    sudo tar -C /usr/local -xzf /tmp/go1.21.3.linux-$(dpkg --print-architecture).tar.gz && \
+    rm /tmp/go1.21.3.linux-$(dpkg --print-architecture).tar.gz

README.md

@@ -40,7 +40,7 @@ PgCat is stable and used in production to serve hundreds of thousands of queries
       </a>
     </td>
     <td>
-      <a href="https://postgresml.org/blog/scaling-postgresml-to-one-million-requests-per-second">
+      <a href="https://postgresml.org/blog/scaling-postgresml-to-1-million-requests-per-second">
         <img src="./images/postgresml.webp" height="70" width="auto">
       </a>
     </td>
@@ -57,7 +57,7 @@ PgCat is stable and used in production to serve hundreds of thousands of queries
       </a>
     </td>
     <td>
-      <a href="https://postgresml.org/blog/scaling-postgresml-to-one-million-requests-per-second">
+      <a href="https://postgresml.org/blog/scaling-postgresml-to-1-million-requests-per-second">
         PostgresML
       </a>
     </td>
@@ -175,7 +175,7 @@ The setting will persist until it's changed again or the client disconnects.
 By default, all queries are routed to the first available server; `default_role` setting controls this behavior.
 
 ### Failover
-All servers are checked with a `;` (very fast) query before being given to a client. Additionally, the server health is monitored with every client query that it processes. If the server is not reachable, it will be banned and cannot serve any more transactions for the duration of the ban. The queries are routed to the remaining servers. If all servers become banned, the ban list is cleared: this is a safety precaution against false positives. The primary can never be banned.
+All servers are checked with a `;` (very fast) query before being given to a client. Additionally, the server health is monitored with every client query that it processes. If the server is not reachable, it will be banned and cannot serve any more transactions for the duration of the ban. The queries are routed to the remaining servers. If `replica_to_primary_failover_enabled` is set to true and all replicas become banned, the query will be routed to the primary. If `replica_to_primary_failover_enabled` is false and all servers (replicas) become banned, the ban list is cleared: this is a safety precaution against false positives. The primary can never be banned.
 
 The ban time can be changed with `ban_time`. The default is 60 seconds.
 
@@ -268,6 +268,8 @@ psql -h 127.0.0.1 -p 6432 -d pgbouncer -c 'SHOW DATABASES'
 
 Additionally, Prometheus statistics are available at `/metrics` via HTTP.
 
+We also have a [basic Grafana dashboard](https://github.com/postgresml/pgcat/blob/main/grafana_dashboard.json) based on Prometheus metrics that you can import into Grafana and build on it or use it for monitoring.
+
 ### Live configuration reloading
 
 The config can be reloaded by sending a `kill -s SIGHUP` to the process or by querying `RELOAD` to the admin database. All settings except the `host` and `port` can be reloaded without restarting the pooler, including sharding and replicas configurations.

charts/pgcat/.helmignore

@@ -0,0 +1,23 @@
+# Patterns to ignore when building packages.
+# This supports shell glob matching, relative path matching, and
+# negation (prefixed with !). Only one pattern per line.
+.DS_Store
+# Common VCS dirs
+.git/
+.gitignore
+.bzr/
+.bzrignore
+.hg/
+.hgignore
+.svn/
+# Common backup files
+*.swp
+*.bak
+*.tmp
+*.orig
+*~
+# Various IDEs
+.project
+.idea/
+*.tmproj
+.vscode/

charts/pgcat/Chart.yaml

@@ -0,0 +1,8 @@
+apiVersion: v2
+name: pgcat
+description: A Helm chart for PgCat a PostgreSQL pooler and proxy (like PgBouncer) with support for sharding, load balancing, failover and mirroring.
+maintainers:
+  - name: Wildcard
+    email: support@w6d.io
+appVersion: "1.2.0"
+version: 0.2.4

charts/pgcat/templates/NOTES.txt

@@ -0,0 +1,22 @@
+1. Get the application URL by running these commands:
+{{- if .Values.ingress.enabled }}
+{{- range $host := .Values.ingress.hosts }}
+  {{- range .paths }}
+  http{{ if $.Values.ingress.tls }}s{{ end }}://{{ $host.host }}{{ .path }}
+  {{- end }}
+{{- end }}
+{{- else if contains "NodePort" .Values.service.type }}
+  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "pgcat.fullname" . }})
+  export NODE_IP=$(kubectl get nodes --namespace {{ .Release.Namespace }} -o jsonpath="{.items[0].status.addresses[0].address}")
+  echo http://$NODE_IP:$NODE_PORT
+{{- else if contains "LoadBalancer" .Values.service.type }}
+     NOTE: It may take a few minutes for the LoadBalancer IP to be available.
+           You can watch the status of by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "pgcat.fullname" . }}'
+  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "pgcat.fullname" . }} --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
+  echo http://$SERVICE_IP:{{ .Values.service.port }}
+{{- else if contains "ClusterIP" .Values.service.type }}
+  export POD_NAME=$(kubectl get pods --namespace {{ .Release.Namespace }} -l "app.kubernetes.io/name={{ include "pgcat.name" . }},app.kubernetes.io/instance={{ .Release.Name }}" -o jsonpath="{.items[0].metadata.name}")
+  export CONTAINER_PORT=$(kubectl get pod --namespace {{ .Release.Namespace }} $POD_NAME -o jsonpath="{.spec.containers[0].ports[0].containerPort}")
+  echo "Visit http://127.0.0.1:8080 to use your application"
+  kubectl --namespace {{ .Release.Namespace }} port-forward $POD_NAME 8080:$CONTAINER_PORT
+{{- end }}

charts/pgcat/templates/_config.tpl

@@ -0,0 +1,3 @@
+{{/*
+  Configuration template definition
+*/}}

charts/pgcat/templates/_helpers.tpl

@@ -0,0 +1,62 @@
+{{/*
+Expand the name of the chart.
+*/}}
+{{- define "pgcat.name" -}}
+{{- default .Chart.Name .Values.nameOverride | trunc 63 | trimSuffix "-" }}
+{{- end }}
+
+{{/*
+Create a default fully qualified app name.
+We truncate at 63 chars because some Kubernetes name fields are limited to this (by the DNS naming spec).
+If release name contains chart name it will be used as a full name.
+*/}}
+{{- define "pgcat.fullname" -}}
+{{- if .Values.fullnameOverride }}
+{{- .Values.fullnameOverride | trunc 63 | trimSuffix "-" }}
+{{- else }}
+{{- $name := default .Chart.Name .Values.nameOverride }}
+{{- if contains $name .Release.Name }}
+{{- .Release.Name | trunc 63 | trimSuffix "-" }}
+{{- else }}
+{{- printf "%s-%s" .Release.Name $name | trunc 63 | trimSuffix "-" }}
+{{- end }}
+{{- end }}
+{{- end }}
+
+{{/*
+Create chart name and version as used by the chart label.
+*/}}
+{{- define "pgcat.chart" -}}
+{{- printf "%s-%s" .Chart.Name .Chart.Version | replace "+" "_" | trunc 63 | trimSuffix "-" }}
+{{- end }}
+
+{{/*
+Common labels
+*/}}
+{{- define "pgcat.labels" -}}
+helm.sh/chart: {{ include "pgcat.chart" . }}
+{{ include "pgcat.selectorLabels" . }}
+{{- if .Chart.AppVersion }}
+app.kubernetes.io/version: {{ .Chart.AppVersion | quote }}
+{{- end }}
+app.kubernetes.io/managed-by: {{ .Release.Service }}
+{{- end }}
+
+{{/*
+Selector labels
+*/}}
+{{- define "pgcat.selectorLabels" -}}
+app.kubernetes.io/name: {{ include "pgcat.name" . }}
+app.kubernetes.io/instance: {{ .Release.Name }}
+{{- end }}
+
+{{/*
+Create the name of the service account to use
+*/}}
+{{- define "pgcat.serviceAccountName" -}}
+{{- if .Values.serviceAccount.create }}
+{{- default (include "pgcat.fullname" .) .Values.serviceAccount.name }}
+{{- else }}
+{{- default "default" .Values.serviceAccount.name }}
+{{- end }}
+{{- end }}

charts/pgcat/templates/deployment.yaml

@@ -0,0 +1,66 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: {{ include "pgcat.fullname" . }}
+  labels:
+    {{- include "pgcat.labels" . | nindent 4 }}
+spec:
+  replicas: {{ .Values.replicaCount }}
+  selector:
+    matchLabels:
+      {{- include "pgcat.selectorLabels" . | nindent 6 }}
+  template:
+    metadata:
+      annotations:
+        checksum/secret: {{ include (print $.Template.BasePath "/secret.yaml") . | sha256sum }}
+      {{- with .Values.podAnnotations }}
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      labels:
+        {{- include "pgcat.selectorLabels" . | nindent 8 }}
+    spec:
+      {{- with .Values.image.pullSecrets }}
+      imagePullSecrets:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      serviceAccountName: {{ include "pgcat.serviceAccountName" . }}
+      securityContext:
+        {{- toYaml .Values.podSecurityContext | nindent 8 }}
+      containers:
+        - name: {{ .Chart.Name }}
+          securityContext:
+            {{- toYaml .Values.containerSecurityContext | nindent 12 }}
+          image: "{{ .Values.image.repository }}:{{ .Values.image.tag | default .Chart.AppVersion }}"
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          ports:
+            - name: pgcat
+              containerPort: {{ .Values.configuration.general.port }}
+              protocol: TCP
+          livenessProbe:
+            tcpSocket:
+              port: pgcat
+          readinessProbe:
+            tcpSocket:
+              port: pgcat
+          resources:
+            {{- toYaml .Values.resources | nindent 12 }}
+          volumeMounts:
+          - mountPath: /etc/pgcat
+            name: config
+      {{- with .Values.nodeSelector }}
+      nodeSelector:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.affinity }}
+      affinity:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.tolerations }}
+      tolerations:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      volumes:
+      - secret:
+          defaultMode: 420
+          secretName: {{ include "pgcat.fullname" . }}
+        name: config

charts/pgcat/templates/ingress.yaml

@@ -0,0 +1,61 @@
+{{- if .Values.ingress.enabled -}}
+{{- $fullName := include "pgcat.fullname" . -}}
+{{- $svcPort := .Values.service.port -}}
+{{- if and .Values.ingress.className (not (semverCompare ">=1.18-0" .Capabilities.KubeVersion.GitVersion)) }}
+  {{- if not (hasKey .Values.ingress.annotations "kubernetes.io/ingress.class") }}
+  {{- $_ := set .Values.ingress.annotations "kubernetes.io/ingress.class" .Values.ingress.className}}
+  {{- end }}
+{{- end }}
+{{- if semverCompare ">=1.19-0" .Capabilities.KubeVersion.GitVersion -}}
+apiVersion: networking.k8s.io/v1
+{{- else if semverCompare ">=1.14-0" .Capabilities.KubeVersion.GitVersion -}}
+apiVersion: networking.k8s.io/v1beta1
+{{- else -}}
+apiVersion: extensions/v1beta1
+{{- end }}
+kind: Ingress
+metadata:
+  name: {{ $fullName }}
+  labels:
+    {{- include "pgcat.labels" . | nindent 4 }}
+  {{- with .Values.ingress.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+spec:
+  {{- if and .Values.ingress.className (semverCompare ">=1.18-0" .Capabilities.KubeVersion.GitVersion) }}
+  ingressClassName: {{ .Values.ingress.className }}
+  {{- end }}
+  {{- if .Values.ingress.tls }}
+  tls:
+    {{- range .Values.ingress.tls }}
+    - hosts:
+        {{- range .hosts }}
+        - {{ . | quote }}
+        {{- end }}
+      secretName: {{ .secretName }}
+    {{- end }}
+  {{- end }}
+  rules:
+    {{- range .Values.ingress.hosts }}
+    - host: {{ .host | quote }}
+      http:
+        paths:
+          {{- range .paths }}
+          - path: {{ .path }}
+            {{- if and .pathType (semverCompare ">=1.18-0" $.Capabilities.KubeVersion.GitVersion) }}
+            pathType: {{ .pathType }}
+            {{- end }}
+            backend:
+              {{- if semverCompare ">=1.19-0" $.Capabilities.KubeVersion.GitVersion }}
+              service:
+                name: {{ $fullName }}
+                port:
+                  number: {{ $svcPort }}
+              {{- else }}
+              serviceName: {{ $fullName }}
+              servicePort: {{ $svcPort }}
+              {{- end }}
+          {{- end }}
+    {{- end }}
+{{- end }}

charts/pgcat/templates/secret.yaml

@@ -0,0 +1,97 @@
+apiVersion: v1
+kind: Secret
+metadata:
+  name: {{ include "pgcat.fullname" . }}
+  labels:
+    {{- include "pgcat.labels" . | nindent 4 }}
+type: Opaque
+stringData:
+  pgcat.toml: |
+    [general]
+    host = {{ .Values.configuration.general.host | quote }}
+    port = {{ .Values.configuration.general.port }}
+    enable_prometheus_exporter = {{ .Values.configuration.general.enable_prometheus_exporter }}
+    prometheus_exporter_port = {{ .Values.configuration.general.prometheus_exporter_port }}
+    connect_timeout = {{ .Values.configuration.general.connect_timeout }}
+    idle_timeout = {{ .Values.configuration.general.idle_timeout | int }}
+    server_lifetime = {{ .Values.configuration.general.server_lifetime | int }}
+    server_tls = {{ .Values.configuration.general.server_tls }}
+    idle_client_in_transaction_timeout = {{ .Values.configuration.general.idle_client_in_transaction_timeout | int }}
+    healthcheck_timeout = {{ .Values.configuration.general.healthcheck_timeout }}
+    healthcheck_delay = {{ .Values.configuration.general.healthcheck_delay }}
+    shutdown_timeout = {{ .Values.configuration.general.shutdown_timeout }}
+    ban_time = {{ .Values.configuration.general.ban_time }}
+    log_client_connections = {{ .Values.configuration.general.log_client_connections }}
+    log_client_disconnections = {{ .Values.configuration.general.log_client_disconnections }}
+    tcp_keepalives_idle = {{ .Values.configuration.general.tcp_keepalives_idle }}
+    tcp_keepalives_count = {{ .Values.configuration.general.tcp_keepalives_count }}
+    tcp_keepalives_interval = {{ .Values.configuration.general.tcp_keepalives_interval }}
+    {{- if and (ne .Values.configuration.general.tls_certificate "-") (ne .Values.configuration.general.tls_private_key "-") }}
+    tls_certificate = "{{ .Values.configuration.general.tls_certificate }}"
+    tls_private_key = "{{ .Values.configuration.general.tls_private_key }}"
+    {{- end }}
+    admin_username = {{ .Values.configuration.general.admin_username | quote }}
+    admin_password = {{ .Values.configuration.general.admin_password | quote }}
+    {{- if and .Values.configuration.general.auth_query_user .Values.configuration.general.auth_query_password .Values.configuration.general.auth_query }}
+    auth_query = {{ .Values.configuration.general.auth_query | quote }}
+    auth_query_user = {{ .Values.configuration.general.auth_query_user | quote }}
+    auth_query_password = {{ .Values.configuration.general.auth_query_password | quote }}
+    {{- end }}
+
+    {{- range $pool := .Values.configuration.pools }}
+
+    ##
+    ## pool for {{ $pool.name }}
+    ##
+    [pools.{{ $pool.name | quote }}]
+    pool_mode = {{ default "transaction" $pool.pool_mode | quote }}
+    load_balancing_mode = {{ default "random" $pool.load_balancing_mode | quote }}
+    default_role = {{ default "any" $pool.default_role | quote }}
+    prepared_statements_cache_size = {{ default 500 $pool.prepared_statements_cache_size }}
+    query_parser_enabled = {{ default true $pool.query_parser_enabled }}
+    query_parser_read_write_splitting = {{ default true $pool.query_parser_read_write_splitting }}
+    primary_reads_enabled = {{ default true $pool.primary_reads_enabled }}
+    sharding_function = {{ default "pg_bigint_hash" $pool.sharding_function | quote }}
+
+    {{-   range $index, $user := $pool.users }}
+
+    ## pool {{ $pool.name }} user {{ $user.username | quote }}
+    ##
+    [pools.{{ $pool.name | quote }}.users.{{ $index }}]
+    username = {{ $user.username | quote }}
+    {{- if $user.password }}
+    password = {{ $user.password | quote }}
+    {{- else if and $user.passwordSecret.name $user.passwordSecret.key }}
+    {{- $secret := (lookup "v1" "Secret" $.Release.Namespace $user.passwordSecret.name) }}
+    {{- if $secret }}
+    {{- $password := index $secret.data $user.passwordSecret.key | b64dec }}
+    password = {{ $password | quote }}
+    {{- end }}
+    {{- end }}
+    pool_size = {{ $user.pool_size }}
+    statement_timeout = {{ default 0 $user.statement_timeout }}
+    min_pool_size = {{ default 3 $user.min_pool_size }}
+    {{- if $user.server_lifetime }}
+    server_lifetime = {{ $user.server_lifetime }}
+    {{- end }}
+    {{-     if and $user.server_username $user.server_password }}
+    server_username = {{ $user.server_username | quote }}
+    server_password = {{ $user.server_password | quote }}
+    {{-     end }}
+    {{-   end }}
+
+    {{-   range $index, $shard := $pool.shards }}
+
+    ## pool {{ $pool.name }} database {{ $shard.database }}
+    ##
+    [pools.{{ $pool.name | quote }}.shards.{{ $index }}]
+    {{-     if gt (len $shard.servers) 0}}
+    servers = [
+    {{-       range  $server := $shard.servers }}
+        [ {{ $server.host | quote }}, {{ $server.port }}, {{ $server.role | quote }} ],
+    {{-       end }}
+    ]
+    {{-     end }}
+    database = {{ $shard.database | quote }}
+    {{-   end }}
+    {{- end }}

charts/pgcat/templates/service.yaml

@@ -0,0 +1,15 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: {{ include "pgcat.fullname" . }}
+  labels:
+    {{- include "pgcat.labels" . | nindent 4 }}
+spec:
+  type: {{ .Values.service.type }}
+  ports:
+    - port: {{ .Values.service.port }}
+      targetPort: pgcat
+      protocol: TCP
+      name: pgcat
+  selector:
+    {{- include "pgcat.selectorLabels" . | nindent 4 }}

charts/pgcat/templates/serviceaccount.yaml

@@ -0,0 +1,12 @@
+{{- if .Values.serviceAccount.create -}}
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ include "pgcat.serviceAccountName" . }}
+  labels:
+    {{- include "pgcat.labels" . | nindent 4 }}
+  {{- with .Values.serviceAccount.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+{{- end }}

charts/pgcat/values.yaml

@@ -0,0 +1,374 @@
+## String to partially override aspnet-core.fullname template (will maintain the release name)
+## @param nameOverride String to partially override common.names.fullname
+##
+nameOverride: ""
+
+## String to fully override aspnet-core.fullname template
+## @param fullnameOverride String to fully override common.names.fullname
+##
+fullnameOverride: ""
+
+## Number of PgCat replicas to deploy
+## @param replicaCount Number of PgCat replicas to deploy
+replicaCount: 1
+
+## Bitnami PgCat image version
+## ref: https://hub.docker.com/r/bitnami/kubewatch/tags/
+##
+## @param image.registry PgCat image registry
+## @param image.repository PgCat image name
+## @param image.tag PgCat image tag
+## @param image.pullPolicy PgCat image tag
+## @param image.pullSecrets Specify docker-registry secret names as an array
+image:
+  repository: ghcr.io/postgresml/pgcat
+  # Overrides the image tag whose default is the chart appVersion.
+  tag: "main"
+  ## Specify a imagePullPolicy
+  ## Defaults to 'Always' if image tag is 'latest', else set to 'IfNotPresent'
+  ## ref: http://kubernetes.io/docs/user-guide/images/#pre-pulling-images
+  ##
+  pullPolicy: IfNotPresent
+  ## Optionally specify an array of imagePullSecrets.
+  ## Secrets must be manually created in the namespace.
+  ## ref: https://kubernetes.io/docs/tasks/configure-pod-container/pull-image-private-registry/
+  ## Example:
+  ## pullSecrets:
+  ##   - myRegistryKeySecretName
+  ##
+  pullSecrets: []
+
+## Specifies whether a ServiceAccount should be created
+##
+## @param serviceAccount.create Enable the creation of a ServiceAccount for PgCat pods
+## @param serviceAccount.name Name of the created ServiceAccount
+##
+serviceAccount:
+  ## Specifies whether a service account should be created
+  create: true
+  ## Annotations to add to the service account
+  annotations: {}
+  ## The name of the service account to use.
+  ## If not set and create is true, a name is generated using the fullname template
+  name: ""
+
+## Annotations for server pods.
+## ref: https://kubernetes.io/docs/concepts/overview/working-with-objects/annotations/
+##
+## @param podAnnotations  Annotations for PgCat pods
+##
+podAnnotations: {}
+
+## PgCat containers' SecurityContext
+## ref: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/#set-the-security-context-for-a-pod
+##
+## @param podSecurityContext.enabled Enabled PgCat pods' Security Context
+## @param podSecurityContext.fsGroup Set PgCat pod's Security Context fsGroup
+##
+podSecurityContext: {}
+  # fsGroup: 2000
+
+## PgCat pods' Security Context
+## ref: https://kubernetes.io/docs/tasks/configure-pod-container/security-context/#set-the-security-context-for-a-container
+##
+## @param containerSecurityContext.enabled Enabled PgCat containers' Security Context
+## @param containerSecurityContext.runAsUser Set PgCat container's Security Context runAsUser
+## @param containerSecurityContext.runAsNonRoot Set PgCat container's Security Context runAsNonRoot
+##
+containerSecurityContext: {}
+  # capabilities:
+  #   drop:
+  #   - ALL
+  # readOnlyRootFilesystem: true
+  # runAsNonRoot: true
+  # runAsUser: 1000
+
+## PgCat service
+##
+## @param service.type PgCat service type
+## @param service.port PgCat service port
+service:
+  type: ClusterIP
+  port: 6432
+
+ingress:
+  enabled: false
+  className: ""
+  annotations: {}
+    # kubernetes.io/ingress.class: nginx
+    # kubernetes.io/tls-acme: "true"
+  hosts:
+    - host: chart-example.local
+      paths:
+        - path: /
+          pathType: ImplementationSpecific
+  tls: []
+  #  - secretName: chart-example-tls
+  #    hosts:
+  #      - chart-example.local
+
+## PgCat resource requests and limits
+## ref: http://kubernetes.io/docs/user-guide/compute-resources/
+##
+## @skip resources Optional description
+## @disabled-param resources.limits  The resources limits for the PgCat container
+## @disabled-param resources.requests  The requested resources for the PgCat container
+##
+resources:
+  # We usually recommend not to specify default resources and to leave this as a conscious
+  # choice for the user. This also increases chances charts run on environments with little
+  # resources, such as Minikube. If you do want to specify resources, uncomment the following
+  # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
+  limits: {}
+  #   cpu: 100m
+  #   memory: 128Mi
+  requests: {}
+  #   cpu: 100m
+  #   memory: 128Mi
+
+## Node labels for pod assignment. Evaluated as a template.
+## ref: https://kubernetes.io/docs/user-guide/node-selection/
+##
+## @param nodeSelector  Node labels for pod assignment
+##
+nodeSelector: {}
+
+## Tolerations for pod assignment. Evaluated as a template.
+## ref: https://kubernetes.io/docs/concepts/configuration/taint-and-toleration/
+##
+## @param tolerations  Tolerations for pod assignment
+##
+tolerations: []
+
+## Affinity for pod assignment. Evaluated as a template.
+## ref: https://kubernetes.io/docs/concepts/configuration/assign-pod-node/#affinity-and-anti-affinity
+## Note: podAffinityPreset, podAntiAffinityPreset, and nodeAffinityPreset will be ignored when it's set
+##
+## @param affinity  Affinity for pod assignment
+##
+affinity: {}
+
+## PgCat configuration
+## @param configuration [object]
+configuration:
+  ## General pooler settings
+  ## @param [object]
+  general:
+    ## @param configuration.general.host What IP to run on, 0.0.0.0 means accessible from everywhere.
+    host: "0.0.0.0"
+
+    ## @param configuration.general.port Port to run on, same as PgBouncer used in this example.
+    port: 6432
+
+    ## @param configuration.general.enable_prometheus_exporter Whether to enable prometheus exporter or not.
+    enable_prometheus_exporter: false
+
+    ## @param configuration.general.prometheus_exporter_port Port at which prometheus exporter listens on.
+    prometheus_exporter_port: 9930
+
+    # @param configuration.general.connect_timeout How long to wait before aborting a server connection (ms).
+    connect_timeout: 5000
+
+    # How long an idle connection with a server is left open (ms).
+    idle_timeout: 30000  # milliseconds
+
+    # Max connection lifetime before it's closed, even if actively used.
+    server_lifetime: 86400000  # 24 hours
+
+    # Whether to use TLS for server connections or not.
+    server_tls: false
+
+    # How long a client is allowed to be idle while in a transaction (ms).
+    idle_client_in_transaction_timeout: 0  # milliseconds
+
+    # @param configuration.general.healthcheck_timeout How much time to give `SELECT 1` health check query to return with a result (ms).
+    healthcheck_timeout: 1000
+
+    # @param configuration.general.healthcheck_delay How long to keep connection available for immediate re-use, without running a healthcheck query on it
+    healthcheck_delay: 30000
+
+    # @param configuration.general.shutdown_timeout How much time to give clients during shutdown before forcibly killing client connections (ms).
+    shutdown_timeout: 60000
+
+    # @param configuration.general.ban_time For how long to ban a server if it fails a health check (seconds).
+    ban_time: 60    # seconds
+
+    # @param configuration.general.log_client_connections If we should log client connections
+    log_client_connections: false
+
+    # @param configuration.general.log_client_disconnections If we should log client disconnections
+    log_client_disconnections: false
+
+    # TLS
+    # tls_certificate: "server.cert"
+    # tls_private_key: "server.key"
+    tls_certificate: "-"
+    tls_private_key: "-"
+
+    # Credentials to access the virtual administrative database (pgbouncer or pgcat)
+    # Connecting to that database allows running commands like `SHOW POOLS`, `SHOW DATABASES`, etc..
+    admin_username: "postgres"
+    admin_password: "postgres"
+
+    # Query to be sent to servers to obtain the hash used for md5 authentication. The connection will be
+    # established using the database configured in the pool. This parameter is inherited by every pool and
+    # can be redefined in pool configuration.
+    auth_query: null
+
+    # User to be used for connecting to servers to obtain the hash used for md5 authentication by sending
+    # the query specified in auth_query_user. The connection will be established using the database configured
+    # in the pool. This parameter is inherited by every pool and can be redefined in pool configuration.
+    #
+    # @param configuration.general.auth_query_user
+    auth_query_user: null
+
+    # Password to be used for connecting to servers to obtain the hash used for md5 authentication by sending
+    # the query specified in auth_query_user. The connection will be established using the database configured
+    # in the pool. This parameter is inherited by every pool and can be redefined in pool configuration.
+    #
+    # @param configuration.general.auth_query_password
+    auth_query_password: null
+
+    # Number of seconds of connection idleness to wait before sending a keepalive packet to the server.
+    tcp_keepalives_idle: 5
+
+    # Number of unacknowledged keepalive packets allowed before giving up and closing the connection.
+    tcp_keepalives_count: 5
+
+    # Number of seconds between keepalive packets.
+    tcp_keepalives_interval: 5
+
+  ## pool
+  ## configs are structured as pool.<pool_name>
+  ## the pool_name is what clients use as database name when connecting
+  ## For the example below a client can connect using "postgres://sharding_user:sharding_user@pgcat_host:pgcat_port/sharded"
+  ## @param [object]
+  pools:
+    [{
+      name: "simple", pool_mode: "transaction",
+      users: [{username: "user", password: "pass", pool_size: 5, statement_timeout: 0}],
+      shards: [{
+        servers: [{host: "postgres", port: 5432, role: "primary"}],
+        database: "postgres"
+      }]
+    }]
+    # - ## default values
+    #   ##
+    #   ##
+    #   ##
+    #   name: "db"
+
+    #  ## Pool mode (see PgBouncer docs for more).
+    #  ## session: one server connection per connected client
+    #  ## transaction: one server connection per client transaction
+    #  ## @param configuration.poolsPostgres.pool_mode
+    #  pool_mode: "transaction"
+
+    #  ## Load balancing mode
+    #  ## `random` selects the server at random
+    #  ## `loc` selects the server with the least outstanding busy connections
+    #  ##
+    #  ## @param configuration.poolsPostgres.load_balancing_mode
+    #  load_balancing_mode: "random"
+
+    #  ## Prepared statements cache size.
+    #  ## TODO: update documentation
+    #  ##
+    #  ## @param configuration.poolsPostgres.prepared_statements_cache_size
+    #  prepared_statements_cache_size: 500
+
+    #  ## If the client doesn't specify, route traffic to
+    #  ## this role by default.
+    #  ##
+    #  ## any: round-robin between primary and replicas,
+    #  ## replica: round-robin between replicas only without touching the primary,
+    #  ## primary: all queries go to the primary unless otherwise specified.
+    #  ## @param configuration.poolsPostgres.default_role
+    #  default_role: "any"
+
+    #  ## Query parser. If enabled, we'll attempt to parse
+    #  ## every incoming query to determine if it's a read or a write.
+    #  ## If it's a read query, we'll direct it to a replica. Otherwise, if it's a write,
+    #  ## we'll direct it to the primary.
+    #  ## @param configuration.poolsPostgres.query_parser_enabled
+    #  query_parser_enabled: true
+
+    #  ## If the query parser is enabled and this setting is enabled, we'll attempt to
+    #  ## infer the role from the query itself.
+    #  ## @param configuration.poolsPostgres.query_parser_read_write_splitting
+    #  query_parser_read_write_splitting: true
+
+    #  ## If the query parser is enabled and this setting is enabled, the primary will be part of the pool of databases used for
+    #  ## load balancing of read queries. Otherwise, the primary will only be used for write
+    #  ## queries. The primary can always be explicitly selected with our custom protocol.
+    #  ## @param configuration.poolsPostgres.primary_reads_enabled
+    #  primary_reads_enabled: true
+
+    #  ## So what if you wanted to implement a different hashing function,
+    #  ## or you've already built one and you want this pooler to use it?
+    #  ##
+    #  ## Current options:
+    #  ##
+    #  ## pg_bigint_hash: PARTITION BY HASH (Postgres hashing function)
+    #  ## sha1: A hashing function based on SHA1
+    #  ##
+    #  ## @param configuration.poolsPostgres.sharding_function
+    #  sharding_function: "pg_bigint_hash"
+
+    #  ## Credentials for users that may connect to this cluster
+    #  ## @param users [array]
+    #  ## @param users[0].username Name of the env var (required)
+    #  ## @param users[0].password Value for the env var (required) leave empty to use existing secret see passwordSecret.name and passwordSecret.key
+    #  ## @param users[0].passwordSecret.name Name of the secret containing the password
+    #  ## @param users[0].passwordSecret.key Key in the secret containing the password
+    #  ## @param users[0].pool_size Maximum number of server connections that can be established for this user
+    #  ## @param users[0].statement_timeout Maximum query duration. Dangerous, but protects against DBs that died in a non-obvious way.
+    #  users: []
+    #    # - username: "user"
+    #    #   password: "pass"
+    #    #
+    #    #  # The maximum number of connection from a single Pgcat process to any database in the cluster
+    #    #  # is the sum of pool_size across all users.
+    #    #  pool_size: 9
+    #    #
+    #    #  # Maximum query duration. Dangerous, but protects against DBs that died in a non-obvious way.
+    #    #  statement_timeout: 0
+    #    #
+    #    #  # PostgreSQL username used to connect to the server.
+    #    #  server_username: "postgres
+    #    #
+    #    #  # PostgreSQL password used to connect to the server.
+    #    #  server_password: "postgres
+
+    #  ## @param shards [array]
+    #  ## @param shards[0].server[0].host Host for this shard
+    #  ## @param shards[0].server[0].port Port for this shard
+    #  ## @param shards[0].server[0].role Role for this shard
+    #  shards: []
+    #      # [ host, port, role ]
+    #    # - servers:
+    #    #     - host: "postgres"
+    #    #       port: 5432
+    #    #       role: "primary"
+    #    #     - host: "postgres"
+    #    #       port: 5432
+    #    #       role: "replica"
+    #    #   database: "postgres"
+    #    #   # [ host, port, role ]
+    #    # - servers:
+    #    #     - host: "postgres"
+    #    #       port: 5432
+    #    #       role: "primary"
+    #    #     - host: "postgres"
+    #    #       port: 5432
+    #    #       role: "replica"
+    #    #   database: "postgres"
+    #    #   # [ host, port, role ]
+    #    # - servers:
+    #    #     - host: "postgres"
+    #    #       port: 5432
+    #    #       role: "primary"
+    #    #     - host: "postgres"
+    #    #       port: 5432
+    #    #       role: "replica"
+    #    #   database: "postgres"

cr.yaml

@@ -0,0 +1,2 @@
+sign: false
+pages_branch: main

ct.yaml

@@ -0,0 +1,5 @@
+remote: origin
+target-branch: main
+chart-dirs:
+  - charts
+

dev/Dockerfile

@@ -1,6 +1,8 @@
-FROM rust:1.70-bullseye
+FROM rust:bullseye
 
 # Dependencies
+COPY --from=sclevine/yj /bin/yj /bin/yj
+RUN /bin/yj -h
 RUN apt-get update -y \
     && apt-get install -y \
     llvm-11 psmisc postgresql-contrib postgresql-client \

pgcat.service

@@ -11,6 +11,7 @@ RestartSec=1
 Environment=RUST_LOG=info
 LimitNOFILE=65536
 ExecStart=/usr/bin/pgcat /etc/pgcat.toml
+ExecReload=/bin/kill -SIGHUP $MAINPID
 
 [Install]
 WantedBy=multi-user.target

pgcat.toml

@@ -60,12 +60,6 @@ tcp_keepalives_count = 5
 # Number of seconds between keepalive packets.
 tcp_keepalives_interval = 5
 
-# Handle prepared statements.
-prepared_statements = true
-
-# Prepared statements server cache size.
-prepared_statements_cache_size = 500
-
 # Path to TLS Certificate file to use for TLS connections
 # tls_certificate = ".circleci/server.cert"
 # Path to TLS private key file to use for TLS connections
@@ -156,6 +150,10 @@ load_balancing_mode = "random"
 # `primary` all queries go to the primary unless otherwise specified.
 default_role = "any"
 
+# Prepared statements cache size.
+# TODO: update documentation
+prepared_statements_cache_size = 500
+
 # If Query Parser is enabled, we'll attempt to parse
 # every incoming query to determine if it's a read or a write.
 # If it's a read query, we'll direct it to a replica. Otherwise, if it's a write,
@@ -181,7 +179,7 @@ primary_reads_enabled = true
 # `random`: picks a shard at random
 # `random_healthy`: picks a shard at random favoring shards with the least number of recent errors
 # `shard_<number>`: e.g. shard_0, shard_4, etc. picks a specific shard, everytime
-# no_shard_specified_behavior = "shard_0"
+# default_shard = "shard_0"
 
 # So what if you wanted to implement a different hashing function,
 # or you've already built one and you want this pooler to use it?
@@ -303,6 +301,8 @@ username = "other_user"
 password = "other_user"
 pool_size = 21
 statement_timeout = 15000
+connect_timeout = 1000
+idle_timeout = 1000
 
 # Shard configs are structured as pool.<pool_name>.shards.<shard_id>
 # Each shard config contains a list of servers that make up the shard

postinst

@@ -7,3 +7,7 @@ systemctl enable pgcat
 if ! id pgcat 2> /dev/null; then
 	useradd -s /usr/bin/false pgcat
 fi
+
+if [ -f /etc/pgcat.toml ]; then
+	systemctl start pgcat
+fi

src/admin.rs

@@ -55,7 +55,12 @@ where
 
     let query_parts: Vec<&str> = query.trim_end_matches(';').split_whitespace().collect();
 
-    match query_parts[0].to_ascii_uppercase().as_str() {
+    match query_parts
+        .first()
+        .unwrap_or(&"")
+        .to_ascii_uppercase()
+        .as_str()
+    {
         "BAN" => {
             trace!("BAN");
             ban(stream, query_parts).await
@@ -84,7 +89,12 @@ where
             trace!("SHUTDOWN");
             shutdown(stream).await
         }
-        "SHOW" => match query_parts[1].to_ascii_uppercase().as_str() {
+        "SHOW" => match query_parts
+            .get(1)
+            .unwrap_or(&"")
+            .to_ascii_uppercase()
+            .as_str()
+        {
             "HELP" => {
                 trace!("SHOW HELP");
                 show_help(stream).await
@@ -283,7 +293,7 @@ where
 {
     let mut res = BytesMut::new();
 
-    let detail_msg = vec![
+    let detail_msg = [
         "",
         "SHOW HELP|CONFIG|DATABASES|POOLS|CLIENTS|SERVERS|USERS|VERSION",
         // "SHOW PEERS|PEER_POOLS", // missing PEERS|PEER_POOLS
@@ -301,7 +311,6 @@ where
         // "KILL <db>",
         // "SUSPEND",
         "SHUTDOWN",
-        // "WAIT_CLOSE [<db>]", // missing
     ];
 
     res.put(notify("Console usage", detail_msg.join("\n\t")));
@@ -691,6 +700,8 @@ where
         ("query_count", DataType::Numeric),
         ("error_count", DataType::Numeric),
         ("age_seconds", DataType::Numeric),
+        ("maxwait", DataType::Numeric),
+        ("maxwait_us", DataType::Numeric),
     ];
 
     let new_map = get_client_stats();
@@ -698,6 +709,7 @@ where
     res.put(row_description(&columns));
 
     for (_, client) in new_map {
+        let max_wait = client.max_wait_time.load(Ordering::Relaxed);
         let row = vec![
             format!("{:#010X}", client.client_id()),
             client.pool_name(),
@@ -711,6 +723,8 @@ where
                 .duration_since(client.connect_time())
                 .as_secs()
                 .to_string(),
+            (max_wait / 1_000_000).to_string(),
+            (max_wait % 1_000_000).to_string(),
         ];
 
         res.put(data_row(&row));
@@ -745,6 +759,7 @@ where
         ("age_seconds", DataType::Numeric),
         ("prepare_cache_hit", DataType::Numeric),
         ("prepare_cache_miss", DataType::Numeric),
+        ("prepare_cache_eviction", DataType::Numeric),
         ("prepare_cache_size", DataType::Numeric),
     ];
 
@@ -777,6 +792,10 @@ where
                 .prepared_miss_count
                 .load(Ordering::Relaxed)
                 .to_string(),
+            server
+                .prepared_eviction_count
+                .load(Ordering::Relaxed)
+                .to_string(),
             server
                 .prepared_cache_size
                 .load(Ordering::Relaxed)
@@ -802,7 +821,7 @@ where
     T: tokio::io::AsyncWrite + std::marker::Unpin,
 {
     let parts: Vec<&str> = match tokens.len() == 2 {
-        true => tokens[1].split(",").map(|part| part.trim()).collect(),
+        true => tokens[1].split(',').map(|part| part.trim()).collect(),
         false => Vec::new(),
     };
 
@@ -865,7 +884,7 @@ where
     T: tokio::io::AsyncWrite + std::marker::Unpin,
 {
     let parts: Vec<&str> = match tokens.len() == 2 {
-        true => tokens[1].split(",").map(|part| part.trim()).collect(),
+        true => tokens[1].split(',').map(|part| part.trim()).collect(),
         false => Vec::new(),
     };
 

src/auth_passthrough.rs

@@ -1,3 +1,4 @@
+use crate::config::AuthType;
 use crate::errors::Error;
 use crate::pool::ConnectionPool;
 use crate::server::Server;
@@ -71,6 +72,7 @@ impl AuthPassthrough {
     pub async fn fetch_hash(&self, address: &crate::config::Address) -> Result<String, Error> {
         let auth_user = crate::config::User {
             username: self.user.clone(),
+            auth_type: AuthType::MD5,
             password: Some(self.password.clone()),
             server_username: None,
             server_password: None,
@@ -79,6 +81,8 @@ impl AuthPassthrough {
             pool_mode: None,
             server_lifetime: None,
             min_pool_size: None,
+            connect_timeout: None,
+            idle_timeout: None,
         };
 
         let user = &address.username;

src/cmd_args.rs

@@ -25,7 +25,7 @@ pub struct Args {
 }
 
 pub fn parse() -> Args {
-    return Args::parse();
+    Args::parse()
 }
 
 #[derive(ValueEnum, Clone, Debug)]

src/config.rs

@@ -1,6 +1,6 @@
 /// Parse the configuration file.
 use arc_swap::ArcSwap;
-use log::{error, info, warn};
+use log::{error, info};
 use once_cell::sync::Lazy;
 use regex::Regex;
 use serde::{Deserializer, Serializer};
@@ -38,12 +38,12 @@ pub enum Role {
     Mirror,
 }
 
-impl ToString for Role {
-    fn to_string(&self) -> String {
-        match *self {
-            Role::Primary => "primary".to_string(),
-            Role::Replica => "replica".to_string(),
-            Role::Mirror => "mirror".to_string(),
+impl std::fmt::Display for Role {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            Role::Primary => write!(f, "primary"),
+            Role::Replica => write!(f, "replica"),
+            Role::Mirror => write!(f, "mirror"),
         }
     }
 }
@@ -116,10 +116,10 @@ impl Default for Address {
             host: String::from("127.0.0.1"),
             port: 5432,
             shard: 0,
-            address_index: 0,
-            replica_number: 0,
             database: String::from("database"),
             role: Role::Replica,
+            replica_number: 0,
+            address_index: 0,
             username: String::from("username"),
             pool_name: String::from("pool_name"),
             mirrors: Vec::new(),
@@ -208,6 +208,9 @@ impl Address {
 pub struct User {
     pub username: String,
     pub password: Option<String>,
+
+    #[serde(default = "User::default_auth_type")]
+    pub auth_type: AuthType,
     pub server_username: Option<String>,
     pub server_password: Option<String>,
     pub pool_size: u32,
@@ -216,6 +219,8 @@ pub struct User {
     pub server_lifetime: Option<u64>,
     #[serde(default)] // 0
     pub statement_timeout: u64,
+    pub connect_timeout: Option<u64>,
+    pub idle_timeout: Option<u64>,
 }
 
 impl Default for User {
@@ -223,6 +228,7 @@ impl Default for User {
         User {
             username: String::from("postgres"),
             password: None,
+            auth_type: AuthType::MD5,
             server_username: None,
             server_password: None,
             pool_size: 15,
@@ -230,24 +236,26 @@ impl Default for User {
             statement_timeout: 0,
             pool_mode: None,
             server_lifetime: None,
+            connect_timeout: None,
+            idle_timeout: None,
         }
     }
 }
 
 impl User {
-    fn validate(&self) -> Result<(), Error> {
-        match self.min_pool_size {
-            Some(min_pool_size) => {
-                if min_pool_size > self.pool_size {
-                    error!(
-                        "min_pool_size of {} cannot be larger than pool_size of {}",
-                        min_pool_size, self.pool_size
-                    );
-                    return Err(Error::BadConfig);
-                }
-            }
+    pub fn default_auth_type() -> AuthType {
+        AuthType::MD5
+    }
 
-            None => (),
+    fn validate(&self) -> Result<(), Error> {
+        if let Some(min_pool_size) = self.min_pool_size {
+            if min_pool_size > self.pool_size {
+                error!(
+                    "min_pool_size of {} cannot be larger than pool_size of {}",
+                    min_pool_size, self.pool_size
+                );
+                return Err(Error::BadConfig);
+            }
         };
 
         Ok(())
@@ -334,6 +342,9 @@ pub struct General {
     pub admin_username: String,
     pub admin_password: String,
 
+    #[serde(default = "General::default_admin_auth_type")]
+    pub admin_auth_type: AuthType,
+
     #[serde(default = "General::default_validate_config")]
     pub validate_config: bool,
 
@@ -341,12 +352,6 @@ pub struct General {
     pub auth_query: Option<String>,
     pub auth_query_user: Option<String>,
     pub auth_query_password: Option<String>,
-
-    #[serde(default)]
-    pub prepared_statements: bool,
-
-    #[serde(default = "General::default_prepared_statements_cache_size")]
-    pub prepared_statements_cache_size: usize,
 }
 
 impl General {
@@ -354,6 +359,10 @@ impl General {
         "0.0.0.0".into()
     }
 
+    pub fn default_admin_auth_type() -> AuthType {
+        AuthType::MD5
+    }
+
     pub fn default_port() -> u16 {
         5432
     }
@@ -428,10 +437,6 @@ impl General {
     pub fn default_server_round_robin() -> bool {
         true
     }
-
-    pub fn default_prepared_statements_cache_size() -> usize {
-        500
-    }
 }
 
 impl Default for General {
@@ -443,35 +448,34 @@ impl Default for General {
             prometheus_exporter_port: 9930,
             connect_timeout: General::default_connect_timeout(),
             idle_timeout: General::default_idle_timeout(),
-            shutdown_timeout: Self::default_shutdown_timeout(),
-            healthcheck_timeout: Self::default_healthcheck_timeout(),
-            healthcheck_delay: Self::default_healthcheck_delay(),
-            ban_time: Self::default_ban_time(),
-            worker_threads: Self::default_worker_threads(),
-            idle_client_in_transaction_timeout: Self::default_idle_client_in_transaction_timeout(),
             tcp_keepalives_idle: Self::default_tcp_keepalives_idle(),
             tcp_keepalives_count: Self::default_tcp_keepalives_count(),
             tcp_keepalives_interval: Self::default_tcp_keepalives_interval(),
             tcp_user_timeout: Self::default_tcp_user_timeout(),
             log_client_connections: false,
             log_client_disconnections: false,
-            autoreload: None,
             dns_cache_enabled: false,
             dns_max_ttl: Self::default_dns_max_ttl(),
+            shutdown_timeout: Self::default_shutdown_timeout(),
+            healthcheck_timeout: Self::default_healthcheck_timeout(),
+            healthcheck_delay: Self::default_healthcheck_delay(),
+            ban_time: Self::default_ban_time(),
+            idle_client_in_transaction_timeout: Self::default_idle_client_in_transaction_timeout(),
+            server_lifetime: Self::default_server_lifetime(),
+            server_round_robin: Self::default_server_round_robin(),
+            worker_threads: Self::default_worker_threads(),
+            autoreload: None,
             tls_certificate: None,
             tls_private_key: None,
             server_tls: false,
             verify_server_certificate: false,
             admin_username: String::from("admin"),
             admin_password: String::from("admin"),
+            admin_auth_type: AuthType::MD5,
+            validate_config: true,
             auth_query: None,
             auth_query_user: None,
             auth_query_password: None,
-            server_lifetime: Self::default_server_lifetime(),
-            server_round_robin: Self::default_server_round_robin(),
-            validate_config: true,
-            prepared_statements: false,
-            prepared_statements_cache_size: 500,
         }
     }
 }
@@ -488,11 +492,20 @@ pub enum PoolMode {
     Session,
 }
 
-impl ToString for PoolMode {
-    fn to_string(&self) -> String {
-        match *self {
-            PoolMode::Transaction => "transaction".to_string(),
-            PoolMode::Session => "session".to_string(),
+#[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Eq, Copy, Hash)]
+pub enum AuthType {
+    #[serde(alias = "trust", alias = "Trust")]
+    Trust,
+
+    #[serde(alias = "md5", alias = "MD5")]
+    MD5,
+}
+
+impl std::fmt::Display for PoolMode {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            PoolMode::Transaction => write!(f, "transaction"),
+            PoolMode::Session => write!(f, "session"),
         }
     }
 }
@@ -505,12 +518,13 @@ pub enum LoadBalancingMode {
     #[serde(alias = "loc", alias = "LOC", alias = "least_outstanding_connections")]
     LeastOutstandingConnections,
 }
-impl ToString for LoadBalancingMode {
-    fn to_string(&self) -> String {
-        match *self {
-            LoadBalancingMode::Random => "random".to_string(),
+
+impl std::fmt::Display for LoadBalancingMode {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            LoadBalancingMode::Random => write!(f, "random"),
             LoadBalancingMode::LeastOutstandingConnections => {
-                "least_outstanding_connections".to_string()
+                write!(f, "least_outstanding_connections")
             }
         }
     }
@@ -527,6 +541,9 @@ pub struct Pool {
     #[serde(default = "Pool::default_default_role")]
     pub default_role: String,
 
+    #[serde(default)] // False
+    pub replica_to_primary_failover_enabled: bool,
+
     #[serde(default)] // False
     pub query_parser_enabled: bool,
 
@@ -572,6 +589,9 @@ pub struct Pool {
     #[serde(default)] // False
     pub log_client_parameter_status_changes: bool,
 
+    #[serde(default = "Pool::default_prepared_statements_cache_size")]
+    pub prepared_statements_cache_size: usize,
+
     pub plugins: Option<Plugins>,
     pub shards: BTreeMap<String, Shard>,
     pub users: BTreeMap<String, User>,
@@ -621,6 +641,10 @@ impl Pool {
         true
     }
 
+    pub fn default_prepared_statements_cache_size() -> usize {
+        0
+    }
+
     pub fn validate(&mut self) -> Result<(), Error> {
         match self.default_role.as_ref() {
             "any" => (),
@@ -677,9 +701,9 @@ impl Pool {
             Some(key) => {
                 // No quotes in the key so we don't have to compare quoted
                 // to unquoted idents.
-                let key = key.replace("\"", "");
+                let key = key.replace('\"', "");
 
-                if key.split(".").count() != 2 {
+                if key.split('.').count() != 2 {
                     error!(
                         "automatic_sharding_key '{}' must be fully qualified, e.g. t.{}`",
                         key, key
@@ -692,17 +716,14 @@ impl Pool {
             None => None,
         };
 
-        match self.default_shard {
-            DefaultShard::Shard(shard_number) => {
-                if shard_number >= self.shards.len() {
-                    error!("Invalid shard {:?}", shard_number);
-                    return Err(Error::BadConfig);
-                }
+        if let DefaultShard::Shard(shard_number) = self.default_shard {
+            if shard_number >= self.shards.len() {
+                error!("Invalid shard {:?}", shard_number);
+                return Err(Error::BadConfig);
             }
-            _ => (),
         }
 
-        for (_, user) in &self.users {
+        for user in self.users.values() {
             user.validate()?;
         }
 
@@ -715,17 +736,17 @@ impl Default for Pool {
         Pool {
             pool_mode: Self::default_pool_mode(),
             load_balancing_mode: Self::default_load_balancing_mode(),
-            shards: BTreeMap::from([(String::from("1"), Shard::default())]),
-            users: BTreeMap::default(),
             default_role: String::from("any"),
+            replica_to_primary_failover_enabled: false,
             query_parser_enabled: false,
             query_parser_max_length: None,
             query_parser_read_write_splitting: false,
             primary_reads_enabled: false,
-            sharding_function: ShardingFunction::PgBigintHash,
-            automatic_sharding_key: None,
             connect_timeout: None,
             idle_timeout: None,
+            server_lifetime: None,
+            sharding_function: ShardingFunction::PgBigintHash,
+            automatic_sharding_key: None,
             sharding_key_regex: None,
             shard_id_regex: None,
             regex_search_limit: Some(1000),
@@ -733,10 +754,12 @@ impl Default for Pool {
             auth_query: None,
             auth_query_user: None,
             auth_query_password: None,
-            server_lifetime: None,
-            plugins: None,
             cleanup_server_connections: true,
             log_client_parameter_status_changes: false,
+            prepared_statements_cache_size: Self::default_prepared_statements_cache_size(),
+            plugins: None,
+            shards: BTreeMap::from([(String::from("1"), Shard::default())]),
+            users: BTreeMap::default(),
         }
     }
 }
@@ -777,8 +800,8 @@ impl<'de> serde::Deserialize<'de> for DefaultShard {
         D: Deserializer<'de>,
     {
         let s = String::deserialize(deserializer)?;
-        if s.starts_with("shard_") {
-            let shard = s[6..].parse::<usize>().map_err(serde::de::Error::custom)?;
+        if let Some(s) = s.strip_prefix("shard_") {
+            let shard = s.parse::<usize>().map_err(serde::de::Error::custom)?;
             return Ok(DefaultShard::Shard(shard));
         }
 
@@ -848,13 +871,13 @@ impl Shard {
 impl Default for Shard {
     fn default() -> Shard {
         Shard {
+            database: String::from("postgres"),
+            mirrors: None,
             servers: vec![ServerConfig {
                 host: String::from("localhost"),
                 port: 5432,
                 role: Role::Primary,
             }],
-            mirrors: None,
-            database: String::from("postgres"),
         }
     }
 }
@@ -867,15 +890,26 @@ pub struct Plugins {
     pub prewarmer: Option<Prewarmer>,
 }
 
+pub trait Plugin {
+    fn is_enabled(&self) -> bool;
+}
+
 impl std::fmt::Display for Plugins {
     fn fmt(&self, f: &mut std::fmt::Formatter) -> std::fmt::Result {
+        fn is_enabled<T: Plugin>(arg: Option<&T>) -> bool {
+            if let Some(arg) = arg {
+                arg.is_enabled()
+            } else {
+                false
+            }
+        }
         write!(
             f,
             "interceptor: {}, table_access: {}, query_logger: {}, prewarmer: {}",
-            self.intercept.is_some(),
-            self.table_access.is_some(),
-            self.query_logger.is_some(),
-            self.prewarmer.is_some(),
+            is_enabled(self.intercept.as_ref()),
+            is_enabled(self.table_access.as_ref()),
+            is_enabled(self.query_logger.as_ref()),
+            is_enabled(self.prewarmer.as_ref()),
         )
     }
 }
@@ -886,23 +920,47 @@ pub struct Intercept {
     pub queries: BTreeMap<String, Query>,
 }
 
+impl Plugin for Intercept {
+    fn is_enabled(&self) -> bool {
+        self.enabled
+    }
+}
+
 #[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Default, Hash, Eq)]
 pub struct TableAccess {
     pub enabled: bool,
     pub tables: Vec<String>,
 }
 
+impl Plugin for TableAccess {
+    fn is_enabled(&self) -> bool {
+        self.enabled
+    }
+}
+
 #[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Default, Hash, Eq)]
 pub struct QueryLogger {
     pub enabled: bool,
 }
 
+impl Plugin for QueryLogger {
+    fn is_enabled(&self) -> bool {
+        self.enabled
+    }
+}
+
 #[derive(Serialize, Deserialize, Debug, Clone, PartialEq, Default, Hash, Eq)]
 pub struct Prewarmer {
     pub enabled: bool,
     pub queries: Vec<String>,
 }
 
+impl Plugin for Prewarmer {
+    fn is_enabled(&self) -> bool {
+        self.enabled
+    }
+}
+
 impl Intercept {
     pub fn substitute(&mut self, db: &str, user: &str) {
         for (_, query) in self.queries.iter_mut() {
@@ -920,6 +978,7 @@ pub struct Query {
 }
 
 impl Query {
+    #[allow(clippy::needless_range_loop)]
     pub fn substitute(&mut self, db: &str, user: &str) {
         for col in self.result.iter_mut() {
             for i in 0..col.len() {
@@ -970,15 +1029,17 @@ impl Config {
     pub fn fill_up_auth_query_config(&mut self) {
         for (_name, pool) in self.pools.iter_mut() {
             if pool.auth_query.is_none() {
-                pool.auth_query = self.general.auth_query.clone();
+                pool.auth_query.clone_from(&self.general.auth_query);
             }
 
             if pool.auth_query_user.is_none() {
-                pool.auth_query_user = self.general.auth_query_user.clone();
+                pool.auth_query_user
+                    .clone_from(&self.general.auth_query_user);
             }
 
             if pool.auth_query_password.is_none() {
-                pool.auth_query_password = self.general.auth_query_password.clone();
+                pool.auth_query_password
+                    .clone_from(&self.general.auth_query_password);
             }
         }
     }
@@ -989,8 +1050,8 @@ impl Default for Config {
         Config {
             path: Self::default_path(),
             general: General::default(),
-            pools: HashMap::default(),
             plugins: None,
+            pools: HashMap::default(),
         }
     }
 }
@@ -1044,8 +1105,8 @@ impl From<&Config> for std::collections::HashMap<String, String> {
                     (
                         format!("pools.{:?}.users", pool_name),
                         pool.users
-                            .iter()
-                            .map(|(_username, user)| &user.username)
+                            .values()
+                            .map(|user| &user.username)
                             .cloned()
                             .collect::<Vec<String>>()
                             .join(", "),
@@ -1099,6 +1160,7 @@ impl From<&Config> for std::collections::HashMap<String, String> {
 impl Config {
     /// Print current configuration.
     pub fn show(&self) {
+        info!("Config path: {}", self.path);
         info!("Ban time: {}s", self.general.ban_time);
         info!(
             "Idle client in transaction timeout: {}ms",
@@ -1125,18 +1187,14 @@ impl Config {
             "Default max server lifetime: {}ms",
             self.general.server_lifetime
         );
-        info!("Sever round robin: {}", self.general.server_round_robin);
+        info!("Server round robin: {}", self.general.server_round_robin);
         match self.general.tls_certificate.clone() {
             Some(tls_certificate) => {
                 info!("TLS certificate: {}", tls_certificate);
 
-                match self.general.tls_private_key.clone() {
-                    Some(tls_private_key) => {
-                        info!("TLS private key: {}", tls_private_key);
-                        info!("TLS support is enabled");
-                    }
-
-                    None => (),
+                if let Some(tls_private_key) = self.general.tls_private_key.clone() {
+                    info!("TLS private key: {}", tls_private_key);
+                    info!("TLS support is enabled");
                 }
             }
 
@@ -1149,13 +1207,6 @@ impl Config {
             "Server TLS certificate verification: {}",
             self.general.verify_server_certificate
         );
-        info!("Prepared statements: {}", self.general.prepared_statements);
-        if self.general.prepared_statements {
-            info!(
-                "Prepared statements server cache size: {}",
-                self.general.prepared_statements_cache_size
-            );
-        }
         info!(
             "Plugins: {}",
             match self.plugins {
@@ -1171,8 +1222,8 @@ impl Config {
                 pool_name,
                 pool_config
                     .users
-                    .iter()
-                    .map(|(_, user_cfg)| user_cfg.pool_size)
+                    .values()
+                    .map(|user_cfg| user_cfg.pool_size)
                     .sum::<u32>()
                     .to_string()
             );
@@ -1246,6 +1297,10 @@ impl Config {
                 "[pool: {}] Log client parameter status changes: {}",
                 pool_name, pool_config.log_client_parameter_status_changes
             );
+            info!(
+                "[pool: {}] Prepared statements server cache size: {}",
+                pool_name, pool_config.prepared_statements_cache_size
+            );
             info!(
                 "[pool: {}] Plugins: {}",
                 pool_name,
@@ -1288,6 +1343,24 @@ impl Config {
                         None => "default".to_string(),
                     }
                 );
+                info!(
+                    "[pool: {}][user: {}] Connection timeout: {}",
+                    pool_name,
+                    user.1.username,
+                    match user.1.connect_timeout {
+                        Some(connect_timeout) => format!("{}ms", connect_timeout),
+                        None => "not set".to_string(),
+                    }
+                );
+                info!(
+                    "[pool: {}][user: {}] Idle timeout: {}",
+                    pool_name,
+                    user.1.username,
+                    match user.1.idle_timeout {
+                        Some(idle_timeout) => format!("{}ms", idle_timeout),
+                        None => "not set".to_string(),
+                    }
+                );
             }
         }
     }
@@ -1342,42 +1415,31 @@ impl Config {
         }
 
         // Validate TLS!
-        match self.general.tls_certificate {
-            Some(ref mut tls_certificate) => {
-                match load_certs(Path::new(&tls_certificate)) {
-                    Ok(_) => {
-                        // Cert is okay, but what about the private key?
-                        match self.general.tls_private_key {
-                            Some(ref tls_private_key) => {
-                                match load_keys(Path::new(&tls_private_key)) {
-                                    Ok(_) => (),
-                                    Err(err) => {
-                                        warn!(
-                                            "tls_private_key is incorrectly configured: {:?}",
-                                            err
-                                        );
-                                        self.general.tls_private_key = None;
-                                        self.general.tls_certificate = None;
-                                    }
-                                }
+        if let Some(tls_certificate) = self.general.tls_certificate.clone() {
+            match load_certs(Path::new(&tls_certificate)) {
+                Ok(_) => {
+                    // Cert is okay, but what about the private key?
+                    match self.general.tls_private_key.clone() {
+                        Some(tls_private_key) => match load_keys(Path::new(&tls_private_key)) {
+                            Ok(_) => (),
+                            Err(err) => {
+                                error!("tls_private_key is incorrectly configured: {:?}", err);
+                                return Err(Error::BadConfig);
                             }
+                        },
 
-                            None => {
-                                warn!("tls_certificate is set, but the tls_private_key is not");
-                                self.general.tls_private_key = None;
-                                self.general.tls_certificate = None;
-                            }
-                        };
-                    }
+                        None => {
+                            error!("tls_certificate is set, but the tls_private_key is not");
+                            return Err(Error::BadConfig);
+                        }
+                    };
+                }
 
-                    Err(err) => {
-                        warn!("tls_certificate is incorrectly configured: {:?}", err);
-                        self.general.tls_private_key = None;
-                        self.general.tls_certificate = None;
-                    }
+                Err(err) => {
+                    error!("tls_certificate is incorrectly configured: {:?}", err);
+                    return Err(Error::BadConfig);
                 }
             }
-            None => (),
         };
 
         for pool in self.pools.values_mut() {
@@ -1399,14 +1461,6 @@ pub fn get_idle_client_in_transaction_timeout() -> u64 {
     CONFIG.load().general.idle_client_in_transaction_timeout
 }
 
-pub fn get_prepared_statements() -> bool {
-    CONFIG.load().general.prepared_statements
-}
-
-pub fn get_prepared_statements_cache_size() -> usize {
-    CONFIG.load().general.prepared_statements_cache_size
-}
-
 /// Parse the configuration file located at the path.
 pub async fn parse(path: &str) -> Result<(), Error> {
     let mut contents = String::new();

src/errors.rs

@@ -29,6 +29,7 @@ pub enum Error {
     QueryRouterParserError(String),
     QueryRouterError(String),
     InvalidShardId(usize),
+    PreparedStatementError,
 }
 
 #[derive(Clone, PartialEq, Debug)]

src/messages.rs

@@ -12,13 +12,16 @@ use crate::config::get_config;
 use crate::errors::Error;
 
 use crate::constants::MESSAGE_TERMINATOR;
+use std::collections::hash_map::DefaultHasher;
 use std::collections::HashMap;
 use std::ffi::CString;
 use std::fmt::{Display, Formatter};
+use std::hash::{Hash, Hasher};
 use std::io::{BufRead, Cursor};
 use std::mem;
 use std::str::FromStr;
 use std::sync::atomic::Ordering;
+use std::sync::Arc;
 use std::time::Duration;
 
 /// Postgres data type mappings
@@ -114,19 +117,11 @@ pub fn simple_query(query: &str) -> BytesMut {
 }
 
 /// Tell the client we're ready for another query.
-pub async fn ready_for_query<S>(stream: &mut S) -> Result<(), Error>
+pub async fn send_ready_for_query<S>(stream: &mut S) -> Result<(), Error>
 where
     S: tokio::io::AsyncWrite + std::marker::Unpin,
 {
-    let mut bytes = BytesMut::with_capacity(
-        mem::size_of::<u8>() + mem::size_of::<i32>() + mem::size_of::<u8>(),
-    );
-
-    bytes.put_u8(b'Z');
-    bytes.put_i32(5);
-    bytes.put_u8(b'I'); // Idle
-
-    write_all(stream, bytes).await
+    write_all(stream, ready_for_query(false)).await
 }
 
 /// Send the startup packet the server. We're pretending we're a Pg client.
@@ -163,12 +158,10 @@ where
 
     match stream.write_all(&startup).await {
         Ok(_) => Ok(()),
-        Err(err) => {
-            return Err(Error::SocketError(format!(
-                "Error writing startup to server socket - Error: {:?}",
-                err
-            )))
-        }
+        Err(err) => Err(Error::SocketError(format!(
+            "Error writing startup to server socket - Error: {:?}",
+            err
+        ))),
     }
 }
 
@@ -244,8 +237,8 @@ pub fn md5_hash_password(user: &str, password: &str, salt: &[u8]) -> Vec<u8> {
     let mut md5 = Md5::new();
 
     // First pass
-    md5.update(&password.as_bytes());
-    md5.update(&user.as_bytes());
+    md5.update(password.as_bytes());
+    md5.update(user.as_bytes());
 
     let output = md5.finalize_reset();
 
@@ -281,7 +274,7 @@ where
 {
     let password = md5_hash_password(user, password, salt);
 
-    let mut message = BytesMut::with_capacity(password.len() as usize + 5);
+    let mut message = BytesMut::with_capacity(password.len() + 5);
 
     message.put_u8(b'p');
     message.put_i32(password.len() as i32 + 4);
@@ -295,7 +288,7 @@ where
     S: tokio::io::AsyncWrite + std::marker::Unpin,
 {
     let password = md5_hash_second_pass(hash, salt);
-    let mut message = BytesMut::with_capacity(password.len() as usize + 5);
+    let mut message = BytesMut::with_capacity(password.len() + 5);
 
     message.put_u8(b'p');
     message.put_i32(password.len() as i32 + 4);
@@ -322,7 +315,7 @@ where
     res.put_slice(&set_complete[..]);
 
     write_all_half(stream, &res).await?;
-    ready_for_query(stream).await
+    send_ready_for_query(stream).await
 }
 
 /// Send a custom error message to the client.
@@ -333,7 +326,7 @@ where
     S: tokio::io::AsyncWrite + std::marker::Unpin,
 {
     error_response_terminal(stream, message).await?;
-    ready_for_query(stream).await
+    send_ready_for_query(stream).await
 }
 
 /// Send a custom error message to the client.
@@ -434,7 +427,7 @@ where
     res.put(command_complete("SELECT 1"));
 
     write_all_half(stream, &res).await?;
-    ready_for_query(stream).await
+    send_ready_for_query(stream).await
 }
 
 pub fn row_description(columns: &Vec<(&str, DataType)>) -> BytesMut {
@@ -516,7 +509,7 @@ pub fn data_row_nullable(row: &Vec<Option<String>>) -> BytesMut {
             data_row.put_i32(column.len() as i32);
             data_row.put_slice(column);
         } else {
-            data_row.put_i32(-1 as i32);
+            data_row.put_i32(-1_i32);
         }
     }
 
@@ -564,6 +557,37 @@ pub fn flush() -> BytesMut {
     bytes
 }
 
+pub fn sync() -> BytesMut {
+    let mut bytes = BytesMut::with_capacity(mem::size_of::<u8>() + mem::size_of::<i32>());
+    bytes.put_u8(b'S');
+    bytes.put_i32(4);
+    bytes
+}
+
+pub fn parse_complete() -> BytesMut {
+    let mut bytes = BytesMut::with_capacity(mem::size_of::<u8>() + mem::size_of::<i32>());
+
+    bytes.put_u8(b'1');
+    bytes.put_i32(4);
+    bytes
+}
+
+pub fn ready_for_query(in_transaction: bool) -> BytesMut {
+    let mut bytes = BytesMut::with_capacity(
+        mem::size_of::<u8>() + mem::size_of::<i32>() + mem::size_of::<u8>(),
+    );
+
+    bytes.put_u8(b'Z');
+    bytes.put_i32(5);
+    if in_transaction {
+        bytes.put_u8(b'T');
+    } else {
+        bytes.put_u8(b'I');
+    }
+
+    bytes
+}
+
 /// Write all data in the buffer to the TcpStream.
 pub async fn write_all<S>(stream: &mut S, buf: BytesMut) -> Result<(), Error>
 where
@@ -571,12 +595,10 @@ where
 {
     match stream.write_all(&buf).await {
         Ok(_) => Ok(()),
-        Err(err) => {
-            return Err(Error::SocketError(format!(
-                "Error writing to socket - Error: {:?}",
-                err
-            )))
-        }
+        Err(err) => Err(Error::SocketError(format!(
+            "Error writing to socket - Error: {:?}",
+            err
+        ))),
     }
 }
 
@@ -587,12 +609,10 @@ where
 {
     match stream.write_all(buf).await {
         Ok(_) => Ok(()),
-        Err(err) => {
-            return Err(Error::SocketError(format!(
-                "Error writing to socket - Error: {:?}",
-                err
-            )))
-        }
+        Err(err) => Err(Error::SocketError(format!(
+            "Error writing to socket - Error: {:?}",
+            err
+        ))),
     }
 }
 
@@ -603,19 +623,15 @@ where
     match stream.write_all(buf).await {
         Ok(_) => match stream.flush().await {
             Ok(_) => Ok(()),
-            Err(err) => {
-                return Err(Error::SocketError(format!(
-                    "Error flushing socket - Error: {:?}",
-                    err
-                )))
-            }
-        },
-        Err(err) => {
-            return Err(Error::SocketError(format!(
-                "Error writing to socket - Error: {:?}",
+            Err(err) => Err(Error::SocketError(format!(
+                "Error flushing socket - Error: {:?}",
                 err
-            )))
-        }
+            ))),
+        },
+        Err(err) => Err(Error::SocketError(format!(
+            "Error writing to socket - Error: {:?}",
+            err
+        ))),
     }
 }
 
@@ -717,6 +733,10 @@ pub fn configure_socket(stream: &TcpStream) {
         }
         Err(err) => error!("Could not configure socket: {}", err),
     }
+    match sock_ref.set_nodelay(true) {
+        Ok(_) => (),
+        Err(err) => error!("Could not configure TCP_NODELAY for socket: {}", err),
+    }
 }
 
 pub trait BytesMutReader {
@@ -730,7 +750,7 @@ impl BytesMutReader for Cursor<&BytesMut> {
         let mut buf = vec![];
         match self.read_until(b'\0', &mut buf) {
             Ok(_) => Ok(String::from_utf8_lossy(&buf[..buf.len() - 1]).to_string()),
-            Err(err) => return Err(Error::ParseBytesError(err.to_string())),
+            Err(err) => Err(Error::ParseBytesError(err.to_string())),
         }
     }
 }
@@ -746,10 +766,55 @@ impl BytesMutReader for BytesMut {
                 let string_bytes = self.split_to(index + 1);
                 Ok(String::from_utf8_lossy(&string_bytes[..string_bytes.len() - 1]).to_string())
             }
-            None => return Err(Error::ParseBytesError("Could not read string".to_string())),
+            None => Err(Error::ParseBytesError("Could not read string".to_string())),
         }
     }
 }
+
+pub enum ExtendedProtocolData {
+    Parse {
+        data: BytesMut,
+        metadata: Option<(Arc<Parse>, u64)>,
+    },
+    Bind {
+        data: BytesMut,
+        metadata: Option<String>,
+    },
+    Describe {
+        data: BytesMut,
+        metadata: Option<String>,
+    },
+    Execute {
+        data: BytesMut,
+    },
+    Close {
+        data: BytesMut,
+        close: Close,
+    },
+}
+
+impl ExtendedProtocolData {
+    pub fn create_new_parse(data: BytesMut, metadata: Option<(Arc<Parse>, u64)>) -> Self {
+        Self::Parse { data, metadata }
+    }
+
+    pub fn create_new_bind(data: BytesMut, metadata: Option<String>) -> Self {
+        Self::Bind { data, metadata }
+    }
+
+    pub fn create_new_describe(data: BytesMut, metadata: Option<String>) -> Self {
+        Self::Describe { data, metadata }
+    }
+
+    pub fn create_new_execute(data: BytesMut) -> Self {
+        Self::Execute { data }
+    }
+
+    pub fn create_new_close(data: BytesMut, close: Close) -> Self {
+        Self::Close { data, close }
+    }
+}
+
 /// Parse (F) message.
 /// See: <https://www.postgresql.org/docs/current/protocol-message-formats.html>
 #[derive(Clone, Debug)]
@@ -758,7 +823,6 @@ pub struct Parse {
     #[allow(dead_code)]
     len: i32,
     pub name: String,
-    pub generated_name: String,
     query: String,
     num_params: i16,
     param_types: Vec<i32>,
@@ -784,7 +848,6 @@ impl TryFrom<&BytesMut> for Parse {
             code,
             len,
             name,
-            generated_name: prepared_statement_name(),
             query,
             num_params,
             param_types,
@@ -833,11 +896,44 @@ impl TryFrom<&Parse> for BytesMut {
 }
 
 impl Parse {
-    pub fn rename(mut self) -> Self {
-        self.name = self.generated_name.to_string();
+    /// Renames the prepared statement to a new name based on the global counter
+    pub fn rewrite(mut self) -> Self {
+        self.name = format!(
+            "PGCAT_{}",
+            PREPARED_STATEMENT_COUNTER.fetch_add(1, Ordering::SeqCst)
+        );
         self
     }
 
+    /// Gets the name of the prepared statement from the buffer
+    pub fn get_name(buf: &BytesMut) -> Result<String, Error> {
+        let mut cursor = Cursor::new(buf);
+        // Skip the code and length
+        cursor.advance(mem::size_of::<u8>() + mem::size_of::<i32>());
+        cursor.read_string()
+    }
+
+    /// Hashes the parse statement to be used as a key in the global cache
+    pub fn get_hash(&self) -> u64 {
+        // TODO_ZAIN: Take a look at which hashing function is being used
+        let mut hasher = DefaultHasher::new();
+
+        let concatenated = format!(
+            "{}{}{}",
+            self.query,
+            self.num_params,
+            self.param_types
+                .iter()
+                .map(ToString::to_string)
+                .collect::<Vec<_>>()
+                .join(",")
+        );
+
+        concatenated.hash(&mut hasher);
+
+        hasher.finish()
+    }
+
     pub fn anonymous(&self) -> bool {
         self.name.is_empty()
     }
@@ -968,9 +1064,42 @@ impl TryFrom<Bind> for BytesMut {
 }
 
 impl Bind {
-    pub fn reassign(mut self, parse: &Parse) -> Self {
-        self.prepared_statement = parse.name.clone();
-        self
+    /// Gets the name of the prepared statement from the buffer
+    pub fn get_name(buf: &BytesMut) -> Result<String, Error> {
+        let mut cursor = Cursor::new(buf);
+        // Skip the code and length
+        cursor.advance(mem::size_of::<u8>() + mem::size_of::<i32>());
+        cursor.read_string()?;
+        cursor.read_string()
+    }
+
+    /// Renames the prepared statement to a new name
+    pub fn rename(buf: BytesMut, new_name: &str) -> Result<BytesMut, Error> {
+        let mut cursor = Cursor::new(&buf);
+        // Read basic data from the cursor
+        let code = cursor.get_u8();
+        let current_len = cursor.get_i32();
+        let portal = cursor.read_string()?;
+        let prepared_statement = cursor.read_string()?;
+
+        // Calculate new length
+        let new_len = current_len + new_name.len() as i32 - prepared_statement.len() as i32;
+
+        // Begin building the response buffer
+        let mut response_buf = BytesMut::with_capacity(new_len as usize + 1);
+        response_buf.put_u8(code);
+        response_buf.put_i32(new_len);
+
+        // Put the portal and new name into the buffer
+        // Note: panic if the provided string contains null byte
+        response_buf.put_slice(CString::new(portal)?.as_bytes_with_nul());
+        response_buf.put_slice(CString::new(new_name)?.as_bytes_with_nul());
+
+        // Add the remainder of the original buffer into the response
+        response_buf.put_slice(&buf[cursor.position() as usize..]);
+
+        // Return the buffer
+        Ok(response_buf)
     }
 
     pub fn anonymous(&self) -> bool {
@@ -984,7 +1113,7 @@ pub struct Describe {
 
     #[allow(dead_code)]
     len: i32,
-    target: char,
+    pub target: char,
     pub statement_name: String,
 }
 
@@ -1026,6 +1155,15 @@ impl TryFrom<Describe> for BytesMut {
 }
 
 impl Describe {
+    pub fn empty_new() -> Describe {
+        Describe {
+            code: 'D',
+            len: 4 + 1 + 1,
+            target: 'S',
+            statement_name: "".to_string(),
+        }
+    }
+
     pub fn rename(mut self, name: &str) -> Self {
         self.statement_name = name.to_string();
         self
@@ -1114,13 +1252,6 @@ pub fn close_complete() -> BytesMut {
     bytes
 }
 
-pub fn prepared_statement_name() -> String {
-    format!(
-        "P_{}",
-        PREPARED_STATEMENT_COUNTER.fetch_add(1, Ordering::SeqCst)
-    )
-}
-
 // from https://www.postgresql.org/docs/12/protocol-error-fields.html
 #[derive(Debug, Default, PartialEq)]
 pub struct PgErrorMsg {
@@ -1203,7 +1334,7 @@ impl Display for PgErrorMsg {
 }
 
 impl PgErrorMsg {
-    pub fn parse(error_msg: Vec<u8>) -> Result<PgErrorMsg, Error> {
+    pub fn parse(error_msg: &[u8]) -> Result<PgErrorMsg, Error> {
         let mut out = PgErrorMsg {
             severity_localized: "".to_string(),
             severity: "".to_string(),
@@ -1311,38 +1442,38 @@ mod tests {
     fn parse_fields() {
         let mut complete_msg = vec![];
         let severity = "FATAL";
-        complete_msg.extend(field('S', &severity));
-        complete_msg.extend(field('V', &severity));
+        complete_msg.extend(field('S', severity));
+        complete_msg.extend(field('V', severity));
 
         let error_code = "29P02";
-        complete_msg.extend(field('C', &error_code));
+        complete_msg.extend(field('C', error_code));
         let message = "password authentication failed for user \"wrong_user\"";
-        complete_msg.extend(field('M', &message));
+        complete_msg.extend(field('M', message));
         let detail_msg = "super detailed message";
-        complete_msg.extend(field('D', &detail_msg));
+        complete_msg.extend(field('D', detail_msg));
         let hint_msg = "hint detail here";
-        complete_msg.extend(field('H', &hint_msg));
+        complete_msg.extend(field('H', hint_msg));
         complete_msg.extend(field('P', "123"));
         complete_msg.extend(field('p', "234"));
         let internal_query = "SELECT * from foo;";
-        complete_msg.extend(field('q', &internal_query));
+        complete_msg.extend(field('q', internal_query));
         let where_msg = "where goes here";
-        complete_msg.extend(field('W', &where_msg));
+        complete_msg.extend(field('W', where_msg));
         let schema_msg = "schema_name";
-        complete_msg.extend(field('s', &schema_msg));
+        complete_msg.extend(field('s', schema_msg));
         let table_msg = "table_name";
-        complete_msg.extend(field('t', &table_msg));
+        complete_msg.extend(field('t', table_msg));
         let column_msg = "column_name";
-        complete_msg.extend(field('c', &column_msg));
+        complete_msg.extend(field('c', column_msg));
         let data_type_msg = "type_name";
-        complete_msg.extend(field('d', &data_type_msg));
+        complete_msg.extend(field('d', data_type_msg));
         let constraint_msg = "constraint_name";
-        complete_msg.extend(field('n', &constraint_msg));
+        complete_msg.extend(field('n', constraint_msg));
         let file_msg = "pgcat.c";
-        complete_msg.extend(field('F', &file_msg));
+        complete_msg.extend(field('F', file_msg));
         complete_msg.extend(field('L', "335"));
         let routine_msg = "my_failing_routine";
-        complete_msg.extend(field('R', &routine_msg));
+        complete_msg.extend(field('R', routine_msg));
 
         tracing_subscriber::fmt()
             .with_max_level(tracing::Level::INFO)
@@ -1351,7 +1482,7 @@ mod tests {
 
         info!(
             "full message: {}",
-            PgErrorMsg::parse(complete_msg.clone()).unwrap()
+            PgErrorMsg::parse(&complete_msg).unwrap()
         );
         assert_eq!(
             PgErrorMsg {
@@ -1374,17 +1505,17 @@ mod tests {
                 line: Some(335),
                 routine: Some(routine_msg.to_string()),
             },
-            PgErrorMsg::parse(complete_msg).unwrap()
+            PgErrorMsg::parse(&complete_msg).unwrap()
         );
 
         let mut only_mandatory_msg = vec![];
-        only_mandatory_msg.extend(field('S', &severity));
-        only_mandatory_msg.extend(field('V', &severity));
-        only_mandatory_msg.extend(field('C', &error_code));
-        only_mandatory_msg.extend(field('M', &message));
-        only_mandatory_msg.extend(field('D', &detail_msg));
+        only_mandatory_msg.extend(field('S', severity));
+        only_mandatory_msg.extend(field('V', severity));
+        only_mandatory_msg.extend(field('C', error_code));
+        only_mandatory_msg.extend(field('M', message));
+        only_mandatory_msg.extend(field('D', detail_msg));
 
-        let err_fields = PgErrorMsg::parse(only_mandatory_msg.clone()).unwrap();
+        let err_fields = PgErrorMsg::parse(&only_mandatory_msg).unwrap();
         info!("only mandatory fields: {}", &err_fields);
         error!(
             "server error: {}: {}",
@@ -1411,7 +1542,7 @@ mod tests {
                 line: None,
                 routine: None,
             },
-            PgErrorMsg::parse(only_mandatory_msg).unwrap()
+            PgErrorMsg::parse(&only_mandatory_msg).unwrap()
         );
     }
 }

src/mirrors.rs

@@ -23,14 +23,15 @@ impl MirroredClient {
     async fn create_pool(&self) -> Pool<ServerPool> {
         let config = get_config();
         let default = std::time::Duration::from_millis(10_000).as_millis() as u64;
-        let (connection_timeout, idle_timeout, _cfg) =
+        let (connection_timeout, idle_timeout, _cfg, prepared_statement_cache_size) =
             match config.pools.get(&self.address.pool_name) {
                 Some(cfg) => (
                     cfg.connect_timeout.unwrap_or(default),
                     cfg.idle_timeout.unwrap_or(default),
                     cfg.clone(),
+                    cfg.prepared_statements_cache_size,
                 ),
-                None => (default, default, crate::config::Pool::default()),
+                None => (default, default, crate::config::Pool::default(), 0),
             };
 
         let manager = ServerPool::new(
@@ -42,6 +43,7 @@ impl MirroredClient {
             None,
             true,
             false,
+            prepared_statement_cache_size,
         );
 
         Pool::builder()
@@ -83,8 +85,9 @@ impl MirroredClient {
                         match recv_result {
                             Ok(message) => trace!("Received from mirror: {} {:?}", String::from_utf8_lossy(&message[..]), address.clone()),
                             Err(err) => {
-                                server.mark_bad();
-                                error!("Failed to receive from mirror {:?} {:?}", err, address.clone());
+                                server.mark_bad(
+                                    format!("Failed to send to mirror, Discarding message {:?}, {:?}", err, address.clone()).as_str()
+                                );
                             }
                         }
                     }
@@ -96,8 +99,9 @@ impl MirroredClient {
                                 match server.send(&BytesMut::from(&bytes[..])).await {
                                     Ok(_) => trace!("Sent to mirror: {} {:?}", String::from_utf8_lossy(&bytes[..]), address.clone()),
                                     Err(err) => {
-                                        server.mark_bad();
-                                        error!("Failed to send to mirror, Discarding message {:?}, {:?}", err, address.clone())
+                                        server.mark_bad(
+                                            format!("Failed to receive from mirror {:?} {:?}", err, address.clone()).as_str()
+                                        );
                                     }
                                 }
                             }
@@ -137,18 +141,18 @@ impl MirroringManager {
                 bytes_rx,
                 disconnect_rx: exit_rx,
             };
-            exit_senders.push(exit_tx.clone());
-            byte_senders.push(bytes_tx.clone());
+            exit_senders.push(exit_tx);
+            byte_senders.push(bytes_tx);
             client.start();
         });
 
         Self {
-            byte_senders: byte_senders,
+            byte_senders,
             disconnect_senders: exit_senders,
         }
     }
 
-    pub fn send(self: &mut Self, bytes: &BytesMut) {
+    pub fn send(&mut self, bytes: &BytesMut) {
         // We want to avoid performing an allocation if we won't be able to send the message
         // There is a possibility of a race here where we check the capacity and then the channel is
         // closed or the capacity is reduced to 0, but mirroring is best effort anyway
@@ -170,7 +174,7 @@ impl MirroringManager {
         });
     }
 
-    pub fn disconnect(self: &mut Self) {
+    pub fn disconnect(&mut self) {
         self.disconnect_senders
             .iter_mut()
             .for_each(|sender| match sender.try_send(()) {

src/plugins/intercept.rs

@@ -92,7 +92,7 @@ impl<'a> Plugin for Intercept<'a> {
                             .map(|s| {
                                 let s = s.as_str().to_string();
 
-                                if s == "" {
+                                if s.is_empty() {
                                     None
                                 } else {
                                     Some(s)

src/plugins/mod.rs

@@ -33,6 +33,7 @@ pub enum PluginOutput {
 #[async_trait]
 pub trait Plugin {
     // Run before the query is sent to the server.
+    #[allow(clippy::ptr_arg)]
     async fn run(
         &mut self,
         query_router: &QueryRouter,

src/plugins/prewarmer.rs

@@ -20,7 +20,7 @@ impl<'a> Prewarmer<'a> {
                 self.server.address(),
                 query
             );
-            self.server.query(&query).await?;
+            self.server.query(query).await?;
         }
 
         Ok(())

src/plugins/table_access.rs

@@ -34,7 +34,7 @@ impl<'a> Plugin for TableAccess<'a> {
 
         visit_relations(ast, |relation| {
             let relation = relation.to_string();
-            let parts = relation.split(".").collect::<Vec<&str>>();
+            let parts = relation.split('.').collect::<Vec<&str>>();
             let table_name = parts.last().unwrap();
 
             if self.tables.contains(&table_name.to_string()) {

src/pool.rs

@@ -3,6 +3,7 @@ use async_trait::async_trait;
 use bb8::{ManageConnection, Pool, PooledConnection, QueueStrategy};
 use chrono::naive::NaiveDateTime;
 use log::{debug, error, info, warn};
+use lru::LruCache;
 use once_cell::sync::Lazy;
 use parking_lot::{Mutex, RwLock};
 use rand::seq::SliceRandom;
@@ -10,6 +11,7 @@ use rand::thread_rng;
 use regex::Regex;
 use std::collections::HashMap;
 use std::fmt::{Display, Formatter};
+use std::num::NonZeroUsize;
 use std::sync::atomic::AtomicU64;
 use std::sync::{
     atomic::{AtomicBool, Ordering},
@@ -24,6 +26,7 @@ use crate::config::{
 use crate::errors::Error;
 
 use crate::auth_passthrough::AuthPassthrough;
+use crate::messages::Parse;
 use crate::plugins::prewarmer;
 use crate::server::{Server, ServerParameters};
 use crate::sharding::ShardingFunction;
@@ -54,6 +57,57 @@ pub enum BanReason {
     AdminBan(i64),
 }
 
+pub type PreparedStatementCacheType = Arc<Mutex<PreparedStatementCache>>;
+
+// TODO: Add stats the this cache
+// TODO: Add application name to the cache value to help identify which application is using the cache
+// TODO: Create admin command to show which statements are in the cache
+#[derive(Debug)]
+pub struct PreparedStatementCache {
+    cache: LruCache<u64, Arc<Parse>>,
+}
+
+impl PreparedStatementCache {
+    pub fn new(mut size: usize) -> Self {
+        // Cannot be zeros
+        if size == 0 {
+            size = 1;
+        }
+
+        PreparedStatementCache {
+            cache: LruCache::new(NonZeroUsize::new(size).unwrap()),
+        }
+    }
+
+    /// Adds the prepared statement to the cache if it doesn't exist with a new name
+    /// if it already exists will give you the existing parse
+    ///
+    /// Pass the hash to this so that we can do the compute before acquiring the lock
+    pub fn get_or_insert(&mut self, parse: &Parse, hash: u64) -> Arc<Parse> {
+        match self.cache.get(&hash) {
+            Some(rewritten_parse) => rewritten_parse.clone(),
+            None => {
+                let new_parse = Arc::new(parse.clone().rewrite());
+                let evicted = self.cache.push(hash, new_parse.clone());
+
+                if let Some((_, evicted_parse)) = evicted {
+                    debug!(
+                        "Evicted prepared statement {} from cache",
+                        evicted_parse.name
+                    );
+                }
+
+                new_parse
+            }
+        }
+    }
+
+    /// Marks the hash as most recently used if it exists
+    pub fn promote(&mut self, hash: &u64) {
+        self.cache.promote(hash);
+    }
+}
+
 /// An identifier for a PgCat pool,
 /// a database visible to clients.
 #[derive(Hash, Debug, Clone, PartialEq, Eq, Default)]
@@ -108,6 +162,9 @@ pub struct PoolSettings {
     // Default server role to connect to.
     pub default_role: Option<Role>,
 
+    // Whether or not we should use primary when replicas are unavailable
+    pub replica_to_primary_failover_enabled: bool,
+
     // Enable/disable query parser.
     pub query_parser_enabled: bool,
 
@@ -165,6 +222,7 @@ impl Default for PoolSettings {
             user: User::default(),
             db: String::default(),
             default_role: None,
+            replica_to_primary_failover_enabled: false,
             query_parser_enabled: false,
             query_parser_max_length: None,
             query_parser_read_write_splitting: false,
@@ -190,11 +248,11 @@ impl Default for PoolSettings {
 #[derive(Clone, Debug, Default)]
 pub struct ConnectionPool {
     /// The pools handled internally by bb8.
-    databases: Vec<Vec<Pool<ServerPool>>>,
+    databases: Arc<Vec<Vec<Pool<ServerPool>>>>,
 
     /// The addresses (host, port, role) to handle
     /// failover and load balancing deterministically.
-    addresses: Vec<Vec<Address>>,
+    addresses: Arc<Vec<Vec<Address>>>,
 
     /// List of banned addresses (see above)
     /// that should not be queried.
@@ -206,7 +264,7 @@ pub struct ConnectionPool {
     original_server_parameters: Arc<RwLock<ServerParameters>>,
 
     /// Pool configuration.
-    pub settings: PoolSettings,
+    pub settings: Arc<PoolSettings>,
 
     /// If not validated, we need to double check the pool is available before allowing a client
     /// to use it.
@@ -223,6 +281,9 @@ pub struct ConnectionPool {
 
     /// AuthInfo
     pub auth_hash: Arc<RwLock<Option<String>>>,
+
+    /// Cache
+    pub prepared_statement_cache: Option<PreparedStatementCacheType>,
 }
 
 impl ConnectionPool {
@@ -241,20 +302,17 @@ impl ConnectionPool {
                 let old_pool_ref = get_pool(pool_name, &user.username);
                 let identifier = PoolIdentifier::new(pool_name, &user.username);
 
-                match old_pool_ref {
-                    Some(pool) => {
-                        // If the pool hasn't changed, get existing reference and insert it into the new_pools.
-                        // We replace all pools at the end, but if the reference is kept, the pool won't get re-created (bb8).
-                        if pool.config_hash == new_pool_hash_value {
-                            info!(
-                                "[pool: {}][user: {}] has not changed",
-                                pool_name, user.username
-                            );
-                            new_pools.insert(identifier.clone(), pool.clone());
-                            continue;
-                        }
+                if let Some(pool) = old_pool_ref {
+                    // If the pool hasn't changed, get existing reference and insert it into the new_pools.
+                    // We replace all pools at the end, but if the reference is kept, the pool won't get re-created (bb8).
+                    if pool.config_hash == new_pool_hash_value {
+                        info!(
+                            "[pool: {}][user: {}] has not changed",
+                            pool_name, user.username
+                        );
+                        new_pools.insert(identifier.clone(), pool.clone());
+                        continue;
                     }
-                    None => (),
                 }
 
                 info!(
@@ -379,16 +437,23 @@ impl ConnectionPool {
                             },
                             pool_config.cleanup_server_connections,
                             pool_config.log_client_parameter_status_changes,
+                            pool_config.prepared_statements_cache_size,
                         );
 
-                        let connect_timeout = match pool_config.connect_timeout {
+                        let connect_timeout = match user.connect_timeout {
                             Some(connect_timeout) => connect_timeout,
-                            None => config.general.connect_timeout,
+                            None => match pool_config.connect_timeout {
+                                Some(connect_timeout) => connect_timeout,
+                                None => config.general.connect_timeout,
+                            },
                         };
 
-                        let idle_timeout = match pool_config.idle_timeout {
+                        let idle_timeout = match user.idle_timeout {
                             Some(idle_timeout) => idle_timeout,
-                            None => config.general.idle_timeout,
+                            None => match pool_config.idle_timeout {
+                                Some(idle_timeout) => idle_timeout,
+                                None => config.general.idle_timeout,
+                            },
                         };
 
                         let server_lifetime = match user.server_lifetime {
@@ -399,7 +464,7 @@ impl ConnectionPool {
                             },
                         };
 
-                        let reaper_rate = *vec![idle_timeout, server_lifetime, POOL_REAPER_RATE]
+                        let reaper_rate = *[idle_timeout, server_lifetime, POOL_REAPER_RATE]
                             .iter()
                             .min()
                             .unwrap();
@@ -448,13 +513,13 @@ impl ConnectionPool {
                 }
 
                 let pool = ConnectionPool {
-                    databases: shards,
-                    addresses,
+                    databases: Arc::new(shards),
+                    addresses: Arc::new(addresses),
                     banlist: Arc::new(RwLock::new(banlist)),
                     config_hash: new_pool_hash_value,
                     original_server_parameters: Arc::new(RwLock::new(ServerParameters::new())),
                     auth_hash: pool_auth_hash,
-                    settings: PoolSettings {
+                    settings: Arc::new(PoolSettings {
                         pool_mode: match user.pool_mode {
                             Some(pool_mode) => pool_mode,
                             None => pool_config.pool_mode,
@@ -470,6 +535,8 @@ impl ConnectionPool {
                             "primary" => Some(Role::Primary),
                             _ => unreachable!(),
                         },
+                        replica_to_primary_failover_enabled: pool_config
+                            .replica_to_primary_failover_enabled,
                         query_parser_enabled: pool_config.query_parser_enabled,
                         query_parser_max_length: pool_config.query_parser_max_length,
                         query_parser_read_write_splitting: pool_config
@@ -489,7 +556,7 @@ impl ConnectionPool {
                             .clone()
                             .map(|regex| Regex::new(regex.as_str()).unwrap()),
                         regex_search_limit: pool_config.regex_search_limit.unwrap_or(1000),
-                        default_shard: pool_config.default_shard.clone(),
+                        default_shard: pool_config.default_shard,
                         auth_query: pool_config.auth_query.clone(),
                         auth_query_user: pool_config.auth_query_user.clone(),
                         auth_query_password: pool_config.auth_query_password.clone(),
@@ -497,17 +564,23 @@ impl ConnectionPool {
                             Some(ref plugins) => Some(plugins.clone()),
                             None => config.plugins.clone(),
                         },
-                    },
+                    }),
                     validated: Arc::new(AtomicBool::new(false)),
                     paused: Arc::new(AtomicBool::new(false)),
                     paused_waiter: Arc::new(Notify::new()),
+                    prepared_statement_cache: match pool_config.prepared_statements_cache_size {
+                        0 => None,
+                        _ => Some(Arc::new(Mutex::new(PreparedStatementCache::new(
+                            pool_config.prepared_statements_cache_size,
+                        )))),
+                    },
                 };
 
                 // Connect to the servers to make sure pool configuration is valid
                 // before setting it globally.
                 // Do this async and somewhere else, we don't have to wait here.
                 if config.general.validate_config {
-                    let mut validate_pool = pool.clone();
+                    let validate_pool = pool.clone();
                     tokio::task::spawn(async move {
                         let _ = validate_pool.validate().await;
                     });
@@ -528,7 +601,7 @@ impl ConnectionPool {
     /// when they connect.
     /// This also warms up the pool for clients that connect when
     /// the pooler starts up.
-    pub async fn validate(&mut self) -> Result<(), Error> {
+    pub async fn validate(&self) -> Result<(), Error> {
         let mut futures = Vec::new();
         let validated = Arc::clone(&self.validated);
 
@@ -664,6 +737,19 @@ impl ConnectionPool {
             });
         }
 
+        // If the role is replica and we allow sending traffic to primary when replicas are unavailble,
+        // we add primary address at the end of the list of candidates, this way it will be tried when
+        // replicas are all unavailable.
+        if role == Role::Replica && self.settings.replica_to_primary_failover_enabled {
+            let mut primaries = self
+                .addresses
+                .iter()
+                .flatten()
+                .filter(|address| address.role == Role::Primary)
+                .collect::<Vec<&Address>>();
+            candidates.insert(0, primaries.pop().unwrap());
+        }
+
         // Indicate we're waiting on a server connection from a pool.
         let now = Instant::now();
         client_stats.waiting();
@@ -678,7 +764,7 @@ impl ConnectionPool {
             let mut force_healthcheck = false;
 
             if self.is_banned(address) {
-                if self.try_unban(&address).await {
+                if self.try_unban(address).await {
                     force_healthcheck = true;
                 } else {
                     debug!("Address {:?} is banned", address);
@@ -702,7 +788,6 @@ impl ConnectionPool {
                     );
                     self.ban(address, BanReason::FailedCheckout, Some(client_stats));
                     address.stats.error();
-                    client_stats.idle();
                     client_stats.checkout_error();
                     continue;
                 }
@@ -721,7 +806,7 @@ impl ConnectionPool {
             // Health checks are pretty expensive.
             if !require_healthcheck {
                 let checkout_time = now.elapsed().as_micros() as u64;
-                client_stats.checkout_time(checkout_time);
+                client_stats.checkout_success();
                 server
                     .stats()
                     .checkout_time(checkout_time, client_stats.application_name());
@@ -735,7 +820,7 @@ impl ConnectionPool {
                 .await
             {
                 let checkout_time = now.elapsed().as_micros() as u64;
-                client_stats.checkout_time(checkout_time);
+                client_stats.checkout_success();
                 server
                     .stats()
                     .checkout_time(checkout_time, client_stats.application_name());
@@ -747,10 +832,7 @@ impl ConnectionPool {
             }
         }
 
-        client_stats.idle();
-
-        let checkout_time = now.elapsed().as_micros() as u64;
-        client_stats.checkout_time(checkout_time);
+        client_stats.checkout_error();
 
         Err(Error::AllServersDown)
     }
@@ -776,7 +858,7 @@ impl ConnectionPool {
             Ok(res) => match res {
                 Ok(_) => {
                     let checkout_time: u64 = start.elapsed().as_micros() as u64;
-                    client_info.checkout_time(checkout_time);
+                    client_info.checkout_success();
                     server
                         .stats()
                         .checkout_time(checkout_time, client_info.application_name());
@@ -804,10 +886,10 @@ impl ConnectionPool {
         }
 
         // Don't leave a bad connection in the pool.
-        server.mark_bad();
+        server.mark_bad("failed health check");
 
-        self.ban(&address, BanReason::FailedHealthCheck, Some(client_info));
-        return false;
+        self.ban(address, BanReason::FailedHealthCheck, Some(client_info));
+        false
     }
 
     /// Ban an address (i.e. replica). It no longer will serve
@@ -872,24 +954,28 @@ impl ConnectionPool {
             return true;
         }
 
-        // Check if all replicas are banned, in that case unban all of them
-        let replicas_available = self.addresses[address.shard]
-            .iter()
-            .filter(|addr| addr.role == Role::Replica)
-            .count();
+        // If we have replica to primary failover we should not unban replicas
+        // as we still have the primary to server traffic.
+        if !self.settings.replica_to_primary_failover_enabled {
+            // Check if all replicas are banned, in that case unban all of them
+            let replicas_available = self.addresses[address.shard]
+                .iter()
+                .filter(|addr| addr.role == Role::Replica)
+                .count();
 
-        debug!("Available targets: {}", replicas_available);
+            debug!("Available targets: {}", replicas_available);
 
-        let read_guard = self.banlist.read();
-        let all_replicas_banned = read_guard[address.shard].len() == replicas_available;
-        drop(read_guard);
+            let read_guard = self.banlist.read();
+            let all_replicas_banned = read_guard[address.shard].len() == replicas_available;
+            drop(read_guard);
 
-        if all_replicas_banned {
-            let mut write_guard = self.banlist.write();
-            warn!("Unbanning all replicas.");
-            write_guard[address.shard].clear();
+            if all_replicas_banned {
+                let mut write_guard = self.banlist.write();
+                warn!("Unbanning all replicas.");
+                write_guard[address.shard].clear();
 
-            return true;
+                return true;
+            }
         }
 
         // Check if ban time is expired
@@ -931,10 +1017,10 @@ impl ConnectionPool {
         let guard = self.banlist.read();
         for banlist in guard.iter() {
             for (address, (reason, timestamp)) in banlist.iter() {
-                bans.push((address.clone(), (reason.clone(), timestamp.clone())));
+                bans.push((address.clone(), (reason.clone(), *timestamp)));
             }
         }
-        return bans;
+        bans
     }
 
     /// Get the address from the host url
@@ -992,7 +1078,7 @@ impl ConnectionPool {
         }
         let busy = provisioned - idle;
         debug!("{:?} has {:?} busy connections", address, busy);
-        return busy;
+        busy
     }
 
     fn valid_shard_id(&self, shard: Option<usize>) -> bool {
@@ -1001,6 +1087,29 @@ impl ConnectionPool {
             Some(shard) => shard < self.shards(),
         }
     }
+
+    /// Register a parse statement to the pool's cache and return the rewritten parse
+    ///
+    /// Do not pass an anonymous parse statement to this function
+    pub fn register_parse_to_cache(&self, hash: u64, parse: &Parse) -> Option<Arc<Parse>> {
+        // We should only be calling this function if the cache is enabled
+        match self.prepared_statement_cache {
+            Some(ref prepared_statement_cache) => {
+                let mut cache = prepared_statement_cache.lock();
+                Some(cache.get_or_insert(parse, hash))
+            }
+            None => None,
+        }
+    }
+
+    /// Promote a prepared statement hash in the LRU
+    pub fn promote_prepared_statement_hash(&self, hash: &u64) {
+        // We should only be calling this function if the cache is enabled
+        if let Some(ref prepared_statement_cache) = self.prepared_statement_cache {
+            let mut cache = prepared_statement_cache.lock();
+            cache.promote(hash);
+        }
+    }
 }
 
 /// Wrapper for the bb8 connection pool.
@@ -1028,9 +1137,13 @@ pub struct ServerPool {
 
     /// Log client parameter status changes
     log_client_parameter_status_changes: bool,
+
+    /// Prepared statement cache size
+    prepared_statement_cache_size: usize,
 }
 
 impl ServerPool {
+    #[allow(clippy::too_many_arguments)]
     pub fn new(
         address: Address,
         user: User,
@@ -1040,16 +1153,18 @@ impl ServerPool {
         plugins: Option<Plugins>,
         cleanup_connections: bool,
         log_client_parameter_status_changes: bool,
+        prepared_statement_cache_size: usize,
     ) -> ServerPool {
         ServerPool {
             address,
-            user: user.clone(),
+            user,
             database: database.to_string(),
             client_server_map,
             auth_hash,
             plugins,
             cleanup_connections,
             log_client_parameter_status_changes,
+            prepared_statement_cache_size,
         }
     }
 }
@@ -1080,6 +1195,7 @@ impl ManageConnection for ServerPool {
             self.auth_hash.clone(),
             self.cleanup_connections,
             self.log_client_parameter_status_changes,
+            self.prepared_statement_cache_size,
         )
         .await
         {

src/prometheus.rs

@@ -1,23 +1,41 @@
-use hyper::service::{make_service_fn, service_fn};
-use hyper::{Body, Method, Request, Response, Server, StatusCode};
+use http_body_util::Full;
+use hyper::body;
+use hyper::body::Bytes;
+
+use hyper::server::conn::http1;
+use hyper::service::service_fn;
+use hyper::{Method, Request, Response, StatusCode};
+use hyper_util::rt::TokioIo;
 use log::{debug, error, info};
 use phf::phf_map;
 use std::collections::HashMap;
 use std::fmt;
 use std::net::SocketAddr;
 use std::sync::atomic::Ordering;
-use std::sync::Arc;
+use tokio::net::TcpListener;
 
 use crate::config::Address;
 use crate::pool::{get_all_pools, PoolIdentifier};
+use crate::stats::get_server_stats;
 use crate::stats::pool::PoolStats;
-use crate::stats::{get_server_stats, ServerStats};
 
 struct MetricHelpType {
     help: &'static str,
     ty: &'static str,
 }
 
+struct ServerPrometheusStats {
+    bytes_received: u64,
+    bytes_sent: u64,
+    transaction_count: u64,
+    query_count: u64,
+    error_count: u64,
+    active_count: u64,
+    idle_count: u64,
+    login_count: u64,
+    tested_count: u64,
+}
+
 // reference for metric types: https://prometheus.io/docs/concepts/metric_types/
 // counters only increase
 // gauges can arbitrarily increase or decrease
@@ -120,22 +138,46 @@ static METRIC_HELP_AND_TYPES_LOOKUP: phf::Map<&'static str, MetricHelpType> = ph
     },
     "servers_bytes_received" => MetricHelpType {
         help: "Volume in bytes of network traffic received by server",
-        ty: "gauge",
+        ty: "counter",
     },
     "servers_bytes_sent" => MetricHelpType {
         help: "Volume in bytes of network traffic sent by server",
-        ty: "gauge",
+        ty: "counter",
     },
     "servers_transaction_count" => MetricHelpType {
         help: "Number of transactions executed by server",
-        ty: "gauge",
+        ty: "counter",
     },
     "servers_query_count" => MetricHelpType {
         help: "Number of queries executed by server",
-        ty: "gauge",
+        ty: "counter",
     },
     "servers_error_count" => MetricHelpType {
         help: "Number of errors",
+        ty: "counter",
+    },
+    "servers_idle_count" => MetricHelpType {
+        help: "Number of server connection in idle state",
+        ty: "gauge",
+    },
+    "servers_active_count" => MetricHelpType {
+        help: "Number of server connection in active state",
+        ty: "gauge",
+    },
+    "servers_tested_count" => MetricHelpType {
+        help: "Number of server connection in tested state",
+        ty: "gauge",
+    },
+    "servers_login_count" => MetricHelpType {
+        help: "Number of server connection in login state",
+        ty: "gauge",
+    },
+    "servers_is_banned" => MetricHelpType {
+        help: "0 if server is not banned, 1 if server is banned",
+        ty: "gauge",
+    },
+    "servers_is_paused" => MetricHelpType {
+        help: "0 if server is not paused, 1 if server is paused",
         ty: "gauge",
     },
     "databases_pool_size" => MetricHelpType {
@@ -158,18 +200,17 @@ struct PrometheusMetric<Value: fmt::Display> {
 
 impl<Value: fmt::Display> fmt::Display for PrometheusMetric<Value> {
     fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
-        let formatted_labels = self
-            .labels
+        let mut sorted_labels: Vec<_> = self.labels.iter().collect();
+        sorted_labels.sort_by_key(|&(key, _)| key);
+        let formatted_labels = sorted_labels
             .iter()
             .map(|(key, value)| format!("{}=\"{}\"", key, value))
             .collect::<Vec<_>>()
             .join(",");
         write!(
             f,
-            "# HELP {name} {help}\n# TYPE {name} {ty}\n{name}{{{formatted_labels}}} {value}\n",
+            "{name}{{{formatted_labels}}} {value}",
             name = format_args!("pgcat_{}", self.name),
-            help = self.help,
-            ty = self.ty,
             formatted_labels = formatted_labels,
             value = self.value
         )
@@ -203,7 +244,9 @@ impl<Value: fmt::Display> PrometheusMetric<Value> {
         labels.insert("shard", address.shard.to_string());
         labels.insert("role", address.role.to_string());
         labels.insert("pool", address.pool_name.clone());
+        labels.insert("index", address.address_index.to_string());
         labels.insert("database", address.database.to_string());
+        labels.insert("username", address.username.clone());
 
         Self::from_name(&format!("databases_{}", name), value, labels)
     }
@@ -218,7 +261,9 @@ impl<Value: fmt::Display> PrometheusMetric<Value> {
         labels.insert("shard", address.shard.to_string());
         labels.insert("role", address.role.to_string());
         labels.insert("pool", address.pool_name.clone());
+        labels.insert("index", address.address_index.to_string());
         labels.insert("database", address.database.to_string());
+        labels.insert("username", address.username.clone());
 
         Self::from_name(&format!("servers_{}", name), value, labels)
     }
@@ -229,7 +274,9 @@ impl<Value: fmt::Display> PrometheusMetric<Value> {
         labels.insert("shard", address.shard.to_string());
         labels.insert("pool", address.pool_name.clone());
         labels.insert("role", address.role.to_string());
+        labels.insert("index", address.address_index.to_string());
         labels.insert("database", address.database.to_string());
+        labels.insert("username", address.username.clone());
 
         Self::from_name(&format!("stats_{}", name), value, labels)
     }
@@ -241,9 +288,20 @@ impl<Value: fmt::Display> PrometheusMetric<Value> {
 
         Self::from_name(&format!("pools_{}", name), value, labels)
     }
+
+    fn get_header(&self) -> String {
+        format!(
+            "\n# HELP {name} {help}\n# TYPE {name} {ty}",
+            name = format_args!("pgcat_{}", self.name),
+            help = self.help,
+            ty = self.ty,
+        )
+    }
 }
 
-async fn prometheus_stats(request: Request<Body>) -> Result<Response<Body>, hyper::http::Error> {
+async fn prometheus_stats(
+    request: Request<body::Incoming>,
+) -> Result<Response<Full<Bytes>>, hyper::http::Error> {
     match (request.method(), request.uri().path()) {
         (&Method::GET, "/metrics") => {
             let mut lines = Vec::new();
@@ -251,6 +309,7 @@ async fn prometheus_stats(request: Request<Body>) -> Result<Response<Body>, hype
             push_pool_stats(&mut lines);
             push_server_stats(&mut lines);
             push_database_stats(&mut lines);
+            lines.push("".to_string()); // Ensure to end the stats with a line terminator as required by the specification.
 
             Response::builder()
                 .header("content-type", "text/plain; version=0.0.4")
@@ -264,6 +323,7 @@ async fn prometheus_stats(request: Request<Body>) -> Result<Response<Body>, hype
 
 // Adds metrics shown in a SHOW STATS admin command.
 fn push_address_stats(lines: &mut Vec<String>) {
+    let mut grouped_metrics: HashMap<String, Vec<PrometheusMetric<u64>>> = HashMap::new();
     for (_, pool) in get_all_pools() {
         for shard in 0..pool.shards() {
             for server in 0..pool.servers(shard) {
@@ -273,7 +333,10 @@ fn push_address_stats(lines: &mut Vec<String>) {
                     if let Some(prometheus_metric) =
                         PrometheusMetric::<u64>::from_address(address, &key, value)
                     {
-                        lines.push(prometheus_metric.to_string());
+                        grouped_metrics
+                            .entry(key)
+                            .or_default()
+                            .push(prometheus_metric);
                     } else {
                         debug!("Metric {} not implemented for {}", key, address.name());
                     }
@@ -281,33 +344,53 @@ fn push_address_stats(lines: &mut Vec<String>) {
             }
         }
     }
+    for (_key, metrics) in grouped_metrics {
+        if !metrics.is_empty() {
+            lines.push(metrics[0].get_header());
+            for metric in metrics {
+                lines.push(metric.to_string());
+            }
+        }
+    }
 }
 
 // Adds relevant metrics shown in a SHOW POOLS admin command.
 fn push_pool_stats(lines: &mut Vec<String>) {
+    let mut grouped_metrics: HashMap<String, Vec<PrometheusMetric<u64>>> = HashMap::new();
     let pool_stats = PoolStats::construct_pool_lookup();
     for (pool_id, stats) in pool_stats.iter() {
         for (name, value) in stats.clone() {
             if let Some(prometheus_metric) =
                 PrometheusMetric::<u64>::from_pool(pool_id.clone(), &name, value)
             {
-                lines.push(prometheus_metric.to_string());
+                grouped_metrics
+                    .entry(name)
+                    .or_default()
+                    .push(prometheus_metric);
             } else {
                 debug!("Metric {} not implemented for ({})", name, *pool_id);
             }
         }
     }
+    for (_key, metrics) in grouped_metrics {
+        if !metrics.is_empty() {
+            lines.push(metrics[0].get_header());
+            for metric in metrics {
+                lines.push(metric.to_string());
+            }
+        }
+    }
 }
 
 // Adds relevant metrics shown in a SHOW DATABASES admin command.
 fn push_database_stats(lines: &mut Vec<String>) {
+    let mut grouped_metrics: HashMap<String, Vec<PrometheusMetric<u32>>> = HashMap::new();
     for (_, pool) in get_all_pools() {
         let pool_config = pool.settings.clone();
         for shard in 0..pool.shards() {
             for server in 0..pool.servers(shard) {
                 let address = pool.address(shard, server);
                 let pool_state = pool.pool_state(shard, server);
-
                 let metrics = vec![
                     ("pool_size", pool_config.user.pool_size),
                     ("current_connections", pool_state.connections),
@@ -316,7 +399,10 @@ fn push_database_stats(lines: &mut Vec<String>) {
                     if let Some(prometheus_metric) =
                         PrometheusMetric::<u32>::from_database_info(address, key, value)
                     {
-                        lines.push(prometheus_metric.to_string());
+                        grouped_metrics
+                            .entry(key.to_string())
+                            .or_default()
+                            .push(prometheus_metric);
                     } else {
                         debug!("Metric {} not implemented for {}", key, address.name());
                     }
@@ -324,45 +410,73 @@ fn push_database_stats(lines: &mut Vec<String>) {
             }
         }
     }
+    for (_key, metrics) in grouped_metrics {
+        if !metrics.is_empty() {
+            lines.push(metrics[0].get_header());
+            for metric in metrics {
+                lines.push(metric.to_string());
+            }
+        }
+    }
 }
 
 // Adds relevant metrics shown in a SHOW SERVERS admin command.
 fn push_server_stats(lines: &mut Vec<String>) {
     let server_stats = get_server_stats();
-    let mut server_stats_by_addresses = HashMap::<String, Arc<ServerStats>>::new();
+    let mut prom_stats = HashMap::<String, ServerPrometheusStats>::new();
     for (_, stats) in server_stats {
-        server_stats_by_addresses.insert(stats.address_name(), stats);
+        let entry = prom_stats
+            .entry(stats.address_name())
+            .or_insert(ServerPrometheusStats {
+                bytes_received: 0,
+                bytes_sent: 0,
+                transaction_count: 0,
+                query_count: 0,
+                error_count: 0,
+                active_count: 0,
+                idle_count: 0,
+                login_count: 0,
+                tested_count: 0,
+            });
+        entry.bytes_received += stats.bytes_received.load(Ordering::Relaxed);
+        entry.bytes_sent += stats.bytes_sent.load(Ordering::Relaxed);
+        entry.transaction_count += stats.transaction_count.load(Ordering::Relaxed);
+        entry.query_count += stats.query_count.load(Ordering::Relaxed);
+        entry.error_count += stats.error_count.load(Ordering::Relaxed);
+        match stats.state.load(Ordering::Relaxed) {
+            crate::stats::ServerState::Login => entry.login_count += 1,
+            crate::stats::ServerState::Active => entry.active_count += 1,
+            crate::stats::ServerState::Tested => entry.tested_count += 1,
+            crate::stats::ServerState::Idle => entry.idle_count += 1,
+        }
     }
-
+    let mut grouped_metrics: HashMap<String, Vec<PrometheusMetric<u64>>> = HashMap::new();
     for (_, pool) in get_all_pools() {
         for shard in 0..pool.shards() {
             for server in 0..pool.servers(shard) {
                 let address = pool.address(shard, server);
-                if let Some(server_info) = server_stats_by_addresses.get(&address.name()) {
+                if let Some(server_info) = prom_stats.get(&address.name()) {
                     let metrics = [
-                        (
-                            "bytes_received",
-                            server_info.bytes_received.load(Ordering::Relaxed),
-                        ),
-                        ("bytes_sent", server_info.bytes_sent.load(Ordering::Relaxed)),
-                        (
-                            "transaction_count",
-                            server_info.transaction_count.load(Ordering::Relaxed),
-                        ),
-                        (
-                            "query_count",
-                            server_info.query_count.load(Ordering::Relaxed),
-                        ),
-                        (
-                            "error_count",
-                            server_info.error_count.load(Ordering::Relaxed),
-                        ),
+                        ("bytes_received", server_info.bytes_received),
+                        ("bytes_sent", server_info.bytes_sent),
+                        ("transaction_count", server_info.transaction_count),
+                        ("query_count", server_info.query_count),
+                        ("error_count", server_info.error_count),
+                        ("idle_count", server_info.idle_count),
+                        ("active_count", server_info.active_count),
+                        ("login_count", server_info.login_count),
+                        ("tested_count", server_info.tested_count),
+                        ("is_banned", if pool.is_banned(address) { 1 } else { 0 }),
+                        ("is_paused", if pool.paused() { 1 } else { 0 }),
                     ];
                     for (key, value) in metrics {
                         if let Some(prometheus_metric) =
                             PrometheusMetric::<u64>::from_server_info(address, key, value)
                         {
-                            lines.push(prometheus_metric.to_string());
+                            grouped_metrics
+                                .entry(key.to_string())
+                                .or_default()
+                                .push(prometheus_metric);
                         } else {
                             debug!("Metric {} not implemented for {}", key, address.name());
                         }
@@ -371,17 +485,46 @@ fn push_server_stats(lines: &mut Vec<String>) {
             }
         }
     }
+    for (_key, metrics) in grouped_metrics {
+        if !metrics.is_empty() {
+            lines.push(metrics[0].get_header());
+            for metric in metrics {
+                lines.push(metric.to_string());
+            }
+        }
+    }
 }
 
 pub async fn start_metric_server(http_addr: SocketAddr) {
-    let http_service_factory =
-        make_service_fn(|_conn| async { Ok::<_, hyper::Error>(service_fn(prometheus_stats)) });
-    let server = Server::bind(&http_addr).serve(http_service_factory);
+    let listener = TcpListener::bind(http_addr);
+    let listener = match listener.await {
+        Ok(listener) => listener,
+        Err(e) => {
+            error!("Failed to bind prometheus server to HTTP address: {}.", e);
+            return;
+        }
+    };
     info!(
         "Exposing prometheus metrics on http://{}/metrics.",
         http_addr
     );
-    if let Err(e) = server.await {
-        error!("Failed to run HTTP server: {}.", e);
+    loop {
+        let stream = match listener.accept().await {
+            Ok((stream, _)) => stream,
+            Err(e) => {
+                error!("Error accepting connection: {}", e);
+                continue;
+            }
+        };
+        let io = TokioIo::new(stream);
+
+        tokio::task::spawn(async move {
+            if let Err(err) = http1::Builder::new()
+                .serve_connection(io, service_fn(prometheus_stats))
+                .await
+            {
+                eprintln!("Error serving HTTP connection for metrics: {:?}", err);
+            }
+        });
     }
 }

src/scram.rs

@@ -79,12 +79,12 @@ impl ScramSha256 {
         let server_message = Message::parse(message)?;
 
         if !server_message.nonce.starts_with(&self.nonce) {
-            return Err(Error::ProtocolSyncError(format!("SCRAM")));
+            return Err(Error::ProtocolSyncError("SCRAM".to_string()));
         }
 
         let salt = match general_purpose::STANDARD.decode(&server_message.salt) {
             Ok(salt) => salt,
-            Err(_) => return Err(Error::ProtocolSyncError(format!("SCRAM"))),
+            Err(_) => return Err(Error::ProtocolSyncError("SCRAM".to_string())),
         };
 
         let salted_password = Self::hi(
@@ -166,9 +166,9 @@ impl ScramSha256 {
     pub fn finish(&mut self, message: &BytesMut) -> Result<(), Error> {
         let final_message = FinalMessage::parse(message)?;
 
-        let verifier = match general_purpose::STANDARD.decode(&final_message.value) {
+        let verifier = match general_purpose::STANDARD.decode(final_message.value) {
             Ok(verifier) => verifier,
-            Err(_) => return Err(Error::ProtocolSyncError(format!("SCRAM"))),
+            Err(_) => return Err(Error::ProtocolSyncError("SCRAM".to_string())),
         };
 
         let mut hmac = match Hmac::<Sha256>::new_from_slice(&self.salted_password) {
@@ -230,14 +230,14 @@ impl Message {
             .collect::<Vec<String>>();
 
         if parts.len() != 3 {
-            return Err(Error::ProtocolSyncError(format!("SCRAM")));
+            return Err(Error::ProtocolSyncError("SCRAM".to_string()));
         }
 
         let nonce = str::replace(&parts[0], "r=", "");
         let salt = str::replace(&parts[1], "s=", "");
         let iterations = match str::replace(&parts[2], "i=", "").parse::<u32>() {
             Ok(iterations) => iterations,
-            Err(_) => return Err(Error::ProtocolSyncError(format!("SCRAM"))),
+            Err(_) => return Err(Error::ProtocolSyncError("SCRAM".to_string())),
         };
 
         Ok(Message {
@@ -257,7 +257,7 @@ impl FinalMessage {
     /// Parse the server final validation message.
     pub fn parse(message: &BytesMut) -> Result<FinalMessage, Error> {
         if !message.starts_with(b"v=") || message.len() < 4 {
-            return Err(Error::ProtocolSyncError(format!("SCRAM")));
+            return Err(Error::ProtocolSyncError("SCRAM".to_string()));
         }
 
         Ok(FinalMessage {

src/server.rs

@@ -3,12 +3,14 @@
 use bytes::{Buf, BufMut, BytesMut};
 use fallible_iterator::FallibleIterator;
 use log::{debug, error, info, trace, warn};
+use lru::LruCache;
 use once_cell::sync::Lazy;
 use parking_lot::{Mutex, RwLock};
 use postgres_protocol::message;
-use std::collections::{BTreeSet, HashMap, HashSet};
+use std::collections::{HashMap, HashSet, VecDeque};
 use std::mem;
 use std::net::IpAddr;
+use std::num::NonZeroUsize;
 use std::sync::Arc;
 use std::time::SystemTime;
 use tokio::io::{AsyncRead, AsyncReadExt, AsyncWrite, BufStream};
@@ -16,7 +18,7 @@ use tokio::net::TcpStream;
 use tokio_rustls::rustls::{OwnedTrustAnchor, RootCertStore};
 use tokio_rustls::{client::TlsStream, TlsConnector};
 
-use crate::config::{get_config, get_prepared_statements_cache_size, Address, User};
+use crate::config::{get_config, Address, User};
 use crate::constants::*;
 use crate::dns_cache::{AddrSet, CACHED_RESOLVER};
 use crate::errors::{Error, ServerIdentifier};
@@ -197,12 +199,8 @@ impl ServerParameters {
             key = "DateStyle".to_string();
         };
 
-        if TRACKED_PARAMETERS.contains(&key) {
+        if TRACKED_PARAMETERS.contains(&key) || startup {
             self.parameters.insert(key, value);
-        } else {
-            if startup {
-                self.parameters.insert(key, value);
-            }
         }
     }
 
@@ -326,12 +324,16 @@ pub struct Server {
     log_client_parameter_status_changes: bool,
 
     /// Prepared statements
-    prepared_statements: BTreeSet<String>,
+    prepared_statement_cache: Option<LruCache<String, ()>>,
+
+    /// Prepared statement being currently registered on the server.
+    registering_prepared_statement: VecDeque<String>,
 }
 
 impl Server {
     /// Pretend to be the Postgres client and connect to the server given host, port and credentials.
     /// Perform the authentication and return the server in a ready for query state.
+    #[allow(clippy::too_many_arguments)]
     pub async fn startup(
         address: &Address,
         user: &User,
@@ -341,6 +343,7 @@ impl Server {
         auth_hash: Arc<RwLock<Option<String>>>,
         cleanup_connections: bool,
         log_client_parameter_status_changes: bool,
+        prepared_statement_cache_size: usize,
     ) -> Result<Server, Error> {
         let cached_resolver = CACHED_RESOLVER.load();
         let mut addr_set: Option<AddrSet> = None;
@@ -440,10 +443,7 @@ impl Server {
 
                 // Something else?
                 m => {
-                    return Err(Error::SocketError(format!(
-                        "Unknown message: {}",
-                        m as char
-                    )));
+                    return Err(Error::SocketError(format!("Unknown message: {}", { m })));
                 }
             }
         } else {
@@ -461,26 +461,20 @@ impl Server {
             None => &user.username,
         };
 
-        let password = match user.server_password {
-            Some(ref server_password) => Some(server_password),
-            None => match user.password {
-                Some(ref password) => Some(password),
-                None => None,
-            },
+        let password = match user.server_password.as_ref() {
+            Some(server_password) => Some(server_password),
+            None => user.password.as_ref(),
         };
 
         startup(&mut stream, username, database).await?;
 
         let mut process_id: i32 = 0;
         let mut secret_key: i32 = 0;
-        let server_identifier = ServerIdentifier::new(username, &database);
+        let server_identifier = ServerIdentifier::new(username, database);
 
         // We'll be handling multiple packets, but they will all be structured the same.
         // We'll loop here until this exchange is complete.
-        let mut scram: Option<ScramSha256> = match password {
-            Some(password) => Some(ScramSha256::new(password)),
-            None => None,
-        };
+        let mut scram: Option<ScramSha256> = password.map(|password| ScramSha256::new(password));
 
         let mut server_parameters = ServerParameters::new();
 
@@ -725,7 +719,7 @@ impl Server {
                                 }
                             };
 
-                            let fields = match PgErrorMsg::parse(error) {
+                            let fields = match PgErrorMsg::parse(&error) {
                                 Ok(f) => f,
                                 Err(err) => {
                                     return Err(err);
@@ -830,7 +824,13 @@ impl Server {
                         },
                         cleanup_connections,
                         log_client_parameter_status_changes,
-                        prepared_statements: BTreeSet::new(),
+                        prepared_statement_cache: match prepared_statement_cache_size {
+                            0 => None,
+                            _ => Some(LruCache::new(
+                                NonZeroUsize::new(prepared_statement_cache_size).unwrap(),
+                            )),
+                        },
+                        registering_prepared_statement: VecDeque::new(),
                     };
 
                     return Ok(server);
@@ -882,7 +882,7 @@ impl Server {
         self.mirror_send(messages);
         self.stats().data_sent(messages.len());
 
-        match write_all_flush(&mut self.stream, &messages).await {
+        match write_all_flush(&mut self.stream, messages).await {
             Ok(_) => {
                 // Successfully sent to server
                 self.last_activity = SystemTime::now();
@@ -960,7 +960,6 @@ impl Server {
 
                     // There is no more data available from the server.
                     self.data_available = false;
-
                     break;
                 }
 
@@ -969,6 +968,37 @@ impl Server {
                     if self.in_copy_mode {
                         self.in_copy_mode = false;
                     }
+
+                    // Remove the prepared statement from the cache, it has a syntax error or something else bad happened.
+                    if let Some(prepared_stmt_name) =
+                        self.registering_prepared_statement.pop_front()
+                    {
+                        if let Some(ref mut cache) = self.prepared_statement_cache {
+                            if let Some(_removed) = cache.pop(&prepared_stmt_name) {
+                                debug!(
+                                    "Removed {} from prepared statement cache",
+                                    prepared_stmt_name
+                                );
+                            } else {
+                                // Shouldn't happen.
+                                debug!("Prepared statement {} was not cached", prepared_stmt_name);
+                            }
+                        }
+                    }
+
+                    if self.prepared_statement_cache.is_some() {
+                        let error_message = PgErrorMsg::parse(&message)?;
+                        if error_message.message == "cached plan must not change result type" {
+                            warn!("Server {:?} changed schema, dropping connection to clean up prepared statements", self.address);
+                            // This will still result in an error to the client, but this server connection will drop all cached prepared statements
+                            // so that any new queries will be re-prepared
+                            // TODO: Other ideas to solve errors when there are DDL changes after a statement has been prepared
+                            //  - Recreate entire connection pool to force recreation of all server connections
+                            //  - Clear the ConnectionPool's statement cache so that new statement names are generated
+                            //  - Implement a retry (re-prepare) so the client doesn't see an error
+                            self.cleanup_state.needs_cleanup_prepare = true;
+                        }
+                    }
                 }
 
                 // CommandComplete
@@ -1058,6 +1088,11 @@ impl Server {
                 // Buffer until ReadyForQuery shows up, so don't exit the loop yet.
                 'c' => (),
 
+                // Parse complete successfully
+                '1' => {
+                    self.registering_prepared_statement.pop_front();
+                }
+
                 // Anything else, e.g. errors, notices, etc.
                 // Keep buffering until ReadyForQuery shows up.
                 _ => (),
@@ -1079,117 +1114,103 @@ impl Server {
         Ok(bytes)
     }
 
-    /// Add the prepared statement to being tracked by this server.
-    /// The client is processing data that will create a prepared statement on this server.
-    pub fn will_prepare(&mut self, name: &str) {
-        debug!("Will prepare `{}`", name);
+    // Determines if the server already has a prepared statement with the given name
+    // Increments the prepared statement cache hit counter
+    pub fn has_prepared_statement(&mut self, name: &str) -> bool {
+        let cache = match &mut self.prepared_statement_cache {
+            Some(cache) => cache,
+            None => return false,
+        };
 
-        self.prepared_statements.insert(name.to_string());
-        self.stats.prepared_cache_add();
-    }
-
-    /// Check if we should prepare a statement on the server.
-    pub fn should_prepare(&self, name: &str) -> bool {
-        let should_prepare = !self.prepared_statements.contains(name);
-
-        debug!("Should prepare `{}`: {}", name, should_prepare);
-
-        if should_prepare {
-            self.stats.prepared_cache_miss();
-        } else {
+        let has_it = cache.get(name).is_some();
+        if has_it {
             self.stats.prepared_cache_hit();
+        } else {
+            self.stats.prepared_cache_miss();
         }
 
-        should_prepare
+        has_it
     }
 
-    /// Create a prepared statement on the server.
-    pub async fn prepare(&mut self, parse: &Parse) -> Result<(), Error> {
-        debug!("Preparing `{}`", parse.name);
+    fn add_prepared_statement_to_cache(&mut self, name: &str) -> Option<String> {
+        let cache = match &mut self.prepared_statement_cache {
+            Some(cache) => cache,
+            None => return None,
+        };
 
-        let bytes: BytesMut = parse.try_into()?;
-        self.send(&bytes).await?;
-        self.send(&flush()).await?;
-
-        // Read and discard ParseComplete (B)
-        match read_message(&mut self.stream).await {
-            Ok(_) => (),
-            Err(err) => {
-                self.bad = true;
-                return Err(err);
-            }
-        }
-
-        self.prepared_statements.insert(parse.name.to_string());
         self.stats.prepared_cache_add();
 
-        debug!("Prepared `{}`", parse.name);
-
-        Ok(())
-    }
-
-    /// Maintain adequate cache size on the server.
-    pub async fn maintain_cache(&mut self) -> Result<(), Error> {
-        debug!("Cache maintenance run");
-
-        let max_cache_size = get_prepared_statements_cache_size();
-        let mut names = Vec::new();
-
-        while self.prepared_statements.len() >= max_cache_size {
-            // The prepared statmeents are alphanumerically sorted by the BTree.
-            // FIFO.
-            if let Some(name) = self.prepared_statements.pop_last() {
-                names.push(name);
+        // If we evict something, we need to close it on the server
+        if let Some((evicted_name, _)) = cache.push(name.to_string(), ()) {
+            if evicted_name != name {
+                debug!(
+                    "Evicted prepared statement {} from cache, replaced with {}",
+                    evicted_name, name
+                );
+                return Some(evicted_name);
             }
-        }
+        };
 
-        if !names.is_empty() {
-            self.deallocate(names).await?;
-        }
-
-        Ok(())
+        None
     }
 
-    /// Remove the prepared statement from being tracked by this server.
-    /// The client is processing data that will cause the server to close the prepared statement.
-    pub fn will_close(&mut self, name: &str) {
-        debug!("Will close `{}`", name);
+    fn remove_prepared_statement_from_cache(&mut self, name: &str) {
+        let cache = match &mut self.prepared_statement_cache {
+            Some(cache) => cache,
+            None => return,
+        };
 
-        self.prepared_statements.remove(name);
+        self.stats.prepared_cache_remove();
+        cache.pop(name);
     }
 
-    /// Close a prepared statement on the server.
-    pub async fn deallocate(&mut self, names: Vec<String>) -> Result<(), Error> {
-        for name in &names {
-            debug!("Deallocating prepared statement `{}`", name);
+    pub async fn register_prepared_statement(
+        &mut self,
+        parse: &Parse,
+        should_send_parse_to_server: bool,
+    ) -> Result<(), Error> {
+        if !self.has_prepared_statement(&parse.name) {
+            self.registering_prepared_statement
+                .push_back(parse.name.clone());
 
-            let close = Close::new(name);
-            let bytes: BytesMut = close.try_into()?;
+            let mut bytes = BytesMut::new();
 
-            self.send(&bytes).await?;
-        }
+            if should_send_parse_to_server {
+                let parse_bytes: BytesMut = parse.try_into()?;
+                bytes.extend_from_slice(&parse_bytes);
+            }
 
-        if !names.is_empty() {
-            self.send(&flush()).await?;
-        }
-
-        // Read and discard CloseComplete (3)
-        for name in &names {
-            match read_message(&mut self.stream).await {
-                Ok(_) => {
-                    self.prepared_statements.remove(name);
-                    self.stats.prepared_cache_remove();
-                    debug!("Closed `{}`", name);
-                }
-
-                Err(err) => {
-                    self.bad = true;
-                    return Err(err);
-                }
+            // If we evict something, we need to close it on the server
+            // We do this by adding it to the messages we're sending to the server before the sync
+            if let Some(evicted_name) = self.add_prepared_statement_to_cache(&parse.name) {
+                self.remove_prepared_statement_from_cache(&evicted_name);
+                let close_bytes: BytesMut = Close::new(&evicted_name).try_into()?;
+                bytes.extend_from_slice(&close_bytes);
             };
-        }
 
-        Ok(())
+            // If we have a parse or close we need to send to the server, send them and sync
+            if !bytes.is_empty() {
+                bytes.extend_from_slice(&sync());
+
+                self.send(&bytes).await?;
+
+                loop {
+                    self.recv(None).await?;
+
+                    if !self.is_data_available() {
+                        break;
+                    }
+                }
+            }
+        };
+
+        // If it's not there, something went bad, I'm guessing bad syntax or permissions error
+        // on the server.
+        if !self.has_prepared_statement(&parse.name) {
+            Err(Error::PreparedStatementError)
+        } else {
+            Ok(())
+        }
     }
 
     /// If the server is still inside a transaction.
@@ -1199,6 +1220,7 @@ impl Server {
         self.in_transaction
     }
 
+    /// Currently copying data from client to server or vice-versa.
     pub fn in_copy_mode(&self) -> bool {
         self.in_copy_mode
     }
@@ -1257,8 +1279,8 @@ impl Server {
     }
 
     /// Indicate that this server connection cannot be re-used and must be discarded.
-    pub fn mark_bad(&mut self) {
-        error!("Server {:?} marked bad", self.address);
+    pub fn mark_bad(&mut self, reason: &str) {
+        error!("Server {:?} marked bad, reason: {}", self.address, reason);
         self.bad = true;
     }
 
@@ -1324,6 +1346,10 @@ impl Server {
 
             if self.cleanup_state.needs_cleanup_prepare {
                 reset_string.push_str("DEALLOCATE ALL;");
+                // Since we deallocated all prepared statements, we need to clear the cache
+                if let Some(cache) = &mut self.prepared_statement_cache {
+                    cache.clear();
+                }
             };
 
             self.query(&reset_string).await?;
@@ -1359,16 +1385,14 @@ impl Server {
     }
 
     pub fn mirror_send(&mut self, bytes: &BytesMut) {
-        match self.mirror_manager.as_mut() {
-            Some(manager) => manager.send(bytes),
-            None => (),
+        if let Some(manager) = self.mirror_manager.as_mut() {
+            manager.send(bytes)
         }
     }
 
     pub fn mirror_disconnect(&mut self) {
-        match self.mirror_manager.as_mut() {
-            Some(manager) => manager.disconnect(),
-            None => (),
+        if let Some(manager) = self.mirror_manager.as_mut() {
+            manager.disconnect()
         }
     }
 
@@ -1391,13 +1415,14 @@ impl Server {
             Arc::new(RwLock::new(None)),
             true,
             false,
+            0,
         )
         .await?;
         debug!("Connected!, sending query.");
         server.send(&simple_query(query)).await?;
         let mut message = server.recv(None).await?;
 
-        Ok(parse_query_message(&mut message).await?)
+        parse_query_message(&mut message).await
     }
 }
 

src/sharding.rs

@@ -14,11 +14,11 @@ pub enum ShardingFunction {
     Sha1,
 }
 
-impl ToString for ShardingFunction {
-    fn to_string(&self) -> String {
-        match *self {
-            ShardingFunction::PgBigintHash => "pg_bigint_hash".to_string(),
-            ShardingFunction::Sha1 => "sha1".to_string(),
+impl std::fmt::Display for ShardingFunction {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        match self {
+            ShardingFunction::PgBigintHash => write!(f, "pg_bigint_hash"),
+            ShardingFunction::Sha1 => write!(f, "sha1"),
         }
     }
 }
@@ -64,7 +64,7 @@ impl Sharder {
     fn sha1(&self, key: i64) -> usize {
         let mut hasher = Sha1::new();
 
-        hasher.update(&key.to_string().as_bytes());
+        hasher.update(key.to_string().as_bytes());
 
         let result = hasher.finalize();
 
@@ -202,10 +202,10 @@ mod test {
     #[test]
     fn test_sha1_hash() {
         let sharder = Sharder::new(12, ShardingFunction::Sha1);
-        let ids = vec![
+        let ids = [
             0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
         ];
-        let shards = vec![
+        let shards = [
             4, 7, 8, 3, 6, 0, 0, 10, 3, 11, 1, 7, 4, 4, 11, 2, 5, 0, 8, 3,
         ];
 

src/stats/client.rs

@@ -41,6 +41,11 @@ pub struct ClientStats {
     /// Maximum time spent waiting for a connection from pool, measures in microseconds
     pub max_wait_time: Arc<AtomicU64>,
 
+    // Time when the client started waiting for a connection from pool, measures in microseconds
+    // We use connect_time as the reference point for this value
+    // U64 can represent ~5850 centuries in microseconds, so we should be fine
+    pub wait_start_us: Arc<AtomicU64>,
+
     /// Current state of the client
     pub state: Arc<AtomicClientState>,
 
@@ -64,6 +69,7 @@ impl Default for ClientStats {
             pool_name: String::new(),
             total_wait_time: Arc::new(AtomicU64::new(0)),
             max_wait_time: Arc::new(AtomicU64::new(0)),
+            wait_start_us: Arc::new(AtomicU64::new(0)),
             state: Arc::new(AtomicClientState::new(ClientState::Idle)),
             transaction_count: Arc::new(AtomicU64::new(0)),
             query_count: Arc::new(AtomicU64::new(0)),
@@ -111,6 +117,9 @@ impl ClientStats {
 
     /// Reports a client is waiting for a connection
     pub fn waiting(&self) {
+        let wait_start = self.connect_time.elapsed().as_micros() as u64;
+
+        self.wait_start_us.store(wait_start, Ordering::Relaxed);
         self.state.store(ClientState::Waiting, Ordering::Relaxed);
     }
 
@@ -122,6 +131,13 @@ impl ClientStats {
     /// Reports a client has failed to obtain a connection from a connection pool
     pub fn checkout_error(&self) {
         self.state.store(ClientState::Idle, Ordering::Relaxed);
+        self.update_wait_times();
+    }
+
+    /// Reports a client has succeeded in obtaining a connection from a connection pool
+    pub fn checkout_success(&self) {
+        self.state.store(ClientState::Active, Ordering::Relaxed);
+        self.update_wait_times();
     }
 
     /// Reports a client has had the server assigned to it be banned
@@ -130,12 +146,26 @@ impl ClientStats {
         self.error_count.fetch_add(1, Ordering::Relaxed);
     }
 
-    /// Reporters the time spent by a client waiting to get a healthy connection from the pool
-    pub fn checkout_time(&self, microseconds: u64) {
+    fn update_wait_times(&self) {
+        if self.wait_start_us.load(Ordering::Relaxed) == 0 {
+            return;
+        }
+
+        let wait_time_us = self.get_current_wait_time_us();
         self.total_wait_time
-            .fetch_add(microseconds, Ordering::Relaxed);
+            .fetch_add(wait_time_us, Ordering::Relaxed);
         self.max_wait_time
-            .fetch_max(microseconds, Ordering::Relaxed);
+            .fetch_max(wait_time_us, Ordering::Relaxed);
+        self.wait_start_us.store(0, Ordering::Relaxed);
+    }
+
+    pub fn get_current_wait_time_us(&self) -> u64 {
+        let wait_start_us = self.wait_start_us.load(Ordering::Relaxed);
+        let microseconds_since_connection_epoch = self.connect_time.elapsed().as_micros() as u64;
+        if wait_start_us == 0 || microseconds_since_connection_epoch < wait_start_us {
+            return 0;
+        }
+        microseconds_since_connection_epoch - wait_start_us
     }
 
     /// Report a query executed by a client against a server

src/stats/pool.rs

@@ -64,8 +64,11 @@ impl PoolStats {
                         ClientState::Idle => pool_stats.cl_idle += 1,
                         ClientState::Waiting => pool_stats.cl_waiting += 1,
                     }
-                    let max_wait = client.max_wait_time.load(Ordering::Relaxed);
-                    pool_stats.maxwait = std::cmp::max(pool_stats.maxwait, max_wait);
+                    let wait_start_us = client.wait_start_us.load(Ordering::Relaxed);
+                    if wait_start_us > 0 {
+                        let wait_time_us = client.get_current_wait_time_us();
+                        pool_stats.maxwait = std::cmp::max(pool_stats.maxwait, wait_time_us);
+                    }
                 }
                 None => debug!("Client from an obselete pool"),
             }
@@ -86,11 +89,11 @@ impl PoolStats {
             }
         }
 
-        return map;
+        map
     }
 
     pub fn generate_header() -> Vec<(&'static str, DataType)> {
-        return vec![
+        vec![
             ("database", DataType::Text),
             ("user", DataType::Text),
             ("pool_mode", DataType::Text),
@@ -105,11 +108,11 @@ impl PoolStats {
             ("sv_login", DataType::Numeric),
             ("maxwait", DataType::Numeric),
             ("maxwait_us", DataType::Numeric),
-        ];
+        ]
     }
 
     pub fn generate_row(&self) -> Vec<String> {
-        return vec![
+        vec![
             self.identifier.db.clone(),
             self.identifier.user.clone(),
             self.mode.to_string(),
@@ -124,7 +127,7 @@ impl PoolStats {
             self.sv_login.to_string(),
             (self.maxwait / 1_000_000).to_string(),
             (self.maxwait % 1_000_000).to_string(),
-        ];
+        ]
     }
 }
 

src/stats/server.rs

@@ -49,6 +49,7 @@ pub struct ServerStats {
     pub error_count: Arc<AtomicU64>,
     pub prepared_hit_count: Arc<AtomicU64>,
     pub prepared_miss_count: Arc<AtomicU64>,
+    pub prepared_eviction_count: Arc<AtomicU64>,
     pub prepared_cache_size: Arc<AtomicU64>,
 }
 
@@ -68,6 +69,7 @@ impl Default for ServerStats {
             reporter: get_reporter(),
             prepared_hit_count: Arc::new(AtomicU64::new(0)),
             prepared_miss_count: Arc::new(AtomicU64::new(0)),
+            prepared_eviction_count: Arc::new(AtomicU64::new(0)),
             prepared_cache_size: Arc::new(AtomicU64::new(0)),
         }
     }
@@ -221,6 +223,7 @@ impl ServerStats {
     }
 
     pub fn prepared_cache_remove(&self) {
+        self.prepared_eviction_count.fetch_add(1, Ordering::Relaxed);
         self.prepared_cache_size.fetch_sub(1, Ordering::Relaxed);
     }
 }

start_test_env.sh

@@ -0,0 +1,34 @@
+GREEN="\033[0;32m"
+RED="\033[0;31m"
+BLUE="\033[0;34m"
+RESET="\033[0m"
+
+
+cd tests/docker/
+docker compose kill main || true
+docker compose build main
+docker compose down
+docker compose up -d
+# wait for the container to start
+while ! docker compose exec main ls; do
+    echo "Waiting for test environment to start"
+    sleep 1
+done
+echo "==================================="
+docker compose exec -e LOG_LEVEL=error -d main toxiproxy-server
+docker compose exec --workdir /app main cargo build
+docker compose exec -d --workdir /app main ./target/debug/pgcat ./.circleci/pgcat.toml
+docker compose exec --workdir /app/tests/ruby main bundle install
+docker compose exec --workdir /app/tests/python main pip3 install -r requirements.txt
+echo "Interactive test environment ready"
+echo "To run integration tests, you can use the following commands:"
+echo -e "   ${BLUE}Ruby:   ${RED}cd /app/tests/ruby && bundle exec ruby tests.rb --format documentation${RESET}"
+echo -e "   ${BLUE}Python: ${RED}cd /app/ && pytest ${RESET}"
+echo -e "   ${BLUE}Rust:   ${RED}cd /app/tests/rust && cargo run ${RESET}"
+echo -e "   ${BLUE}Go:     ${RED}cd /app/tests/go && /usr/local/go/bin/go test${RESET}"
+echo "the source code for tests are directly linked to the source code in the container so you can modify the code and run the tests again"
+echo "You can rebuild PgCat from within the container by running"
+echo -e "  ${GREEN}cargo build${RESET}"
+echo "and then run the tests again"
+echo "==================================="
+docker compose exec --workdir /app/tests main bash

tests/docker/Dockerfile

@@ -8,3 +8,6 @@ RUN rustup component add llvm-tools-preview
 RUN sudo gem install bundler
 RUN wget -O toxiproxy-2.4.0.deb https://github.com/Shopify/toxiproxy/releases/download/v2.4.0/toxiproxy_2.4.0_linux_$(dpkg --print-architecture).deb && \
 	sudo dpkg -i toxiproxy-2.4.0.deb
+RUN wget -O go1.21.3.linux-$(dpkg --print-architecture).tar.gz https://go.dev/dl/go1.21.3.linux-$(dpkg --print-architecture).tar.gz && \
+    sudo tar -C /usr/local -xzf go1.21.3.linux-$(dpkg --print-architecture).tar.gz && \
+    rm go1.21.3.linux-$(dpkg --print-architecture).tar.gz

tests/docker/docker-compose.yml

@@ -1,4 +1,3 @@
-version: "3"
 services:
   pg1:
     image: postgres:14
@@ -48,6 +47,8 @@ services:
   main:
     build: .
     command: ["bash", "/app/tests/docker/run.sh"]
+    environment:
+      - INTERACTIVE_TEST_ENVIRONMENT=true
     volumes:
       - ../../:/app/
       - /app/target/

tests/docker/run.sh

@@ -5,6 +5,38 @@ rm /app/*.profraw || true
 rm /app/pgcat.profdata || true
 rm -rf /app/cov || true
 
+# Prepares the interactive test environment
+# 
+if [ -n "$INTERACTIVE_TEST_ENVIRONMENT" ]; then
+    ports=(5432 7432 8432 9432 10432)
+    for port in "${ports[@]}"; do
+        is_it_up=0
+        attempts=0
+        while [ $is_it_up -eq 0 ]; do
+            PGPASSWORD=postgres psql -h 127.0.0.1 -p $port -U postgres -c '\q' > /dev/null 2>&1
+            if [ $? -eq 0 ]; then
+                echo "PostgreSQL on port $port is up."
+                is_it_up=1
+            else
+                attempts=$((attempts+1))
+                if [ $attempts -gt 10 ]; then
+                    echo "PostgreSQL on port $port is down, giving up."
+                    exit 1
+                fi
+                echo "PostgreSQL on port $port is down, waiting for it to start."
+                sleep 1
+            fi
+        done
+    done
+    PGPASSWORD=postgres psql -e -h 127.0.0.1 -p 5432  -U postgres -f /app/tests/sharding/query_routing_setup.sql
+    PGPASSWORD=postgres psql -e -h 127.0.0.1 -p 7432  -U postgres -f /app/tests/sharding/query_routing_setup.sql
+    PGPASSWORD=postgres psql -e -h 127.0.0.1 -p 8432  -U postgres -f /app/tests/sharding/query_routing_setup.sql
+    PGPASSWORD=postgres psql -e -h 127.0.0.1 -p 9432  -U postgres -f /app/tests/sharding/query_routing_setup.sql
+    PGPASSWORD=postgres psql -e -h 127.0.0.1 -p 10432 -U postgres -f /app/tests/sharding/query_routing_setup.sql
+    sleep 100000000000000000
+    exit 0
+fi
+
 export LLVM_PROFILE_FILE="/app/pgcat-%m-%p.profraw"
 export RUSTC_BOOTSTRAP=1
 export CARGO_INCREMENTAL=0

tests/go/go.mod

@@ -0,0 +1,5 @@
+module pgcat
+
+go 1.21
+
+require github.com/lib/pq v1.10.9

tests/go/go.sum

@@ -0,0 +1,2 @@
+github.com/lib/pq v1.10.9 h1:YXG7RB+JIjhP29X+OtkiDnYaXQwpS4JEWq7dtCCRUEw=
+github.com/lib/pq v1.10.9/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=

tests/go/pgcat.toml

@@ -0,0 +1,162 @@
+#
+# PgCat config example.
+#
+
+#
+# General pooler settings
+[general]
+# What IP to run on, 0.0.0.0 means accessible from everywhere.
+host = "0.0.0.0"
+
+# Port to run on, same as PgBouncer used in this example.
+port = "${PORT}"
+
+# Whether to enable prometheus exporter or not.
+enable_prometheus_exporter = true
+
+# Port at which prometheus exporter listens on.
+prometheus_exporter_port = 9930
+
+# How long to wait before aborting a server connection (ms).
+connect_timeout = 1000
+
+# How much time to give the health check query to return with a result (ms).
+healthcheck_timeout = 1000
+
+# How long to keep connection available for immediate re-use, without running a healthcheck query on it
+healthcheck_delay = 30000
+
+# How much time to give clients during shutdown before forcibly killing client connections (ms).
+shutdown_timeout = 5000
+
+# For how long to ban a server if it fails a health check (seconds).
+ban_time = 60 # Seconds
+
+# If we should log client connections
+log_client_connections = false
+
+# If we should log client disconnections
+log_client_disconnections = false
+
+# Reload config automatically if it changes.
+autoreload = 15000
+
+server_round_robin = false
+
+# TLS
+tls_certificate = "../../.circleci/server.cert"
+tls_private_key = "../../.circleci/server.key"
+
+# Credentials to access the virtual administrative database (pgbouncer or pgcat)
+# Connecting to that database allows running commands like `SHOW POOLS`, `SHOW DATABASES`, etc..
+admin_username = "admin_user"
+admin_password = "admin_pass"
+
+# pool
+# configs are structured as pool.<pool_name>
+# the pool_name is what clients use as database name when connecting
+# For the example below a client can connect using "postgres://sharding_user:sharding_user@pgcat_host:pgcat_port/sharded_db"
+[pools.sharded_db]
+# Pool mode (see PgBouncer docs for more).
+# session: one server connection per connected client
+# transaction: one server connection per client transaction
+pool_mode = "transaction"
+
+# If the client doesn't specify, route traffic to
+# this role by default.
+#
+# any: round-robin between primary and replicas,
+# replica: round-robin between replicas only without touching the primary,
+# primary: all queries go to the primary unless otherwise specified.
+default_role = "any"
+
+# Query parser. If enabled, we'll attempt to parse
+# every incoming query to determine if it's a read or a write.
+# If it's a read query, we'll direct it to a replica. Otherwise, if it's a write,
+# we'll direct it to the primary.
+query_parser_enabled = true
+
+# If the query parser is enabled and this setting is enabled, we'll attempt to
+# infer the role from the query itself.
+query_parser_read_write_splitting = true
+
+# If the query parser is enabled and this setting is enabled, the primary will be part of the pool of databases used for
+# load balancing of read queries. Otherwise, the primary will only be used for write
+# queries. The primary can always be explicitely selected with our custom protocol.
+primary_reads_enabled = true
+
+# So what if you wanted to implement a different hashing function,
+# or you've already built one and you want this pooler to use it?
+#
+# Current options:
+#
+# pg_bigint_hash: PARTITION BY HASH (Postgres hashing function)
+# sha1: A hashing function based on SHA1
+#
+sharding_function = "pg_bigint_hash"
+
+# Prepared statements cache size.
+prepared_statements_cache_size = 500
+
+# Credentials for users that may connect to this cluster
+[pools.sharded_db.users.0]
+username = "sharding_user"
+password = "sharding_user"
+# Maximum number of server connections that can be established for this user
+# The maximum number of connection from a single Pgcat process to any database in the cluster
+# is the sum of pool_size across all users.
+pool_size = 5
+statement_timeout = 0
+
+
+[pools.sharded_db.users.1]
+username = "other_user"
+password = "other_user"
+pool_size = 21
+statement_timeout = 30000
+
+# Shard 0
+[pools.sharded_db.shards.0]
+# [ host, port, role ]
+servers = [
+    [ "127.0.0.1", 5432, "primary" ],
+    [ "localhost", 5432, "replica" ]
+]
+# Database name (e.g. "postgres")
+database = "shard0"
+
+[pools.sharded_db.shards.1]
+servers = [
+    [ "127.0.0.1", 5432, "primary" ],
+    [ "localhost", 5432, "replica" ],
+]
+database = "shard1"
+
+[pools.sharded_db.shards.2]
+servers = [
+    [ "127.0.0.1", 5432, "primary" ],
+    [ "localhost", 5432, "replica" ],
+]
+database = "shard2"
+
+
+[pools.simple_db]
+pool_mode = "session"
+default_role = "primary"
+query_parser_enabled = true
+query_parser_read_write_splitting = true
+primary_reads_enabled = true
+sharding_function = "pg_bigint_hash"
+
+[pools.simple_db.users.0]
+username = "simple_user"
+password = "simple_user"
+pool_size = 5
+statement_timeout = 30000
+
+[pools.simple_db.shards.0]
+servers = [
+    [ "127.0.0.1", 5432, "primary" ],
+    [ "localhost", 5432, "replica" ]
+]
+database = "some_db"

tests/go/prepared_test.go

@@ -0,0 +1,52 @@
+package pgcat
+
+import (
+	"context"
+	"database/sql"
+	"fmt"
+	_ "github.com/lib/pq"
+	"testing"
+)
+
+func Test(t *testing.T) {
+	t.Cleanup(setup(t))
+	t.Run("Named parameterized prepared statement works", namedParameterizedPreparedStatement)
+	t.Run("Unnamed parameterized prepared statement works", unnamedParameterizedPreparedStatement)
+}
+
+func namedParameterizedPreparedStatement(t *testing.T) {
+	db, err := sql.Open("postgres", fmt.Sprintf("host=localhost port=%d database=sharded_db user=sharding_user password=sharding_user sslmode=disable", port))
+	if err != nil {
+		t.Fatalf("could not open connection: %+v", err)
+	}
+
+	stmt, err := db.Prepare("SELECT $1")
+
+	if err != nil {
+		t.Fatalf("could not prepare: %+v", err)
+	}
+
+	for i := 0; i < 100; i++ {
+		rows, err := stmt.Query(1)
+		if err != nil {
+			t.Fatalf("could not query: %+v", err)
+		}
+		_ = rows.Close()
+	}
+}
+
+func unnamedParameterizedPreparedStatement(t *testing.T) {
+	db, err := sql.Open("postgres", fmt.Sprintf("host=localhost port=%d database=sharded_db user=sharding_user password=sharding_user sslmode=disable", port))
+	if err != nil {
+		t.Fatalf("could not open connection: %+v", err)
+	}
+
+	for i := 0; i < 100; i++ {
+		// Under the hood QueryContext generates an unnamed parameterized prepared statement
+		rows, err := db.QueryContext(context.Background(), "SELECT $1", 1)
+		if err != nil {
+			t.Fatalf("could not query: %+v", err)
+		}
+		_ = rows.Close()
+	}
+}

tests/go/setup.go

@@ -0,0 +1,81 @@
+package pgcat
+
+import (
+	"context"
+	"database/sql"
+	_ "embed"
+	"fmt"
+	"math/rand"
+	"os"
+	"os/exec"
+	"strings"
+	"testing"
+	"time"
+)
+
+//go:embed pgcat.toml
+var pgcatCfg string
+
+var port = rand.Intn(32760-20000) + 20000
+
+func setup(t *testing.T) func() {
+	cfg, err := os.CreateTemp("/tmp", "pgcat_cfg_*.toml")
+	if err != nil {
+		t.Fatalf("could not create temp file: %+v", err)
+	}
+
+	pgcatCfg = strings.Replace(pgcatCfg, "\"${PORT}\"", fmt.Sprintf("%d", port), 1)
+
+	_, err = cfg.Write([]byte(pgcatCfg))
+	if err != nil {
+		t.Fatalf("could not write temp file: %+v", err)
+	}
+
+	commandPath := "../../target/debug/pgcat"
+	if os.Getenv("CARGO_TARGET_DIR") != "" {
+		commandPath = os.Getenv("CARGO_TARGET_DIR") + "/debug/pgcat"
+	}
+
+	cmd := exec.Command(commandPath, cfg.Name())
+	cmd.Stdout = os.Stdout
+	cmd.Stderr = os.Stderr
+	go func() {
+		err = cmd.Run()
+		if err != nil {
+			t.Errorf("could not run pgcat: %+v", err)
+		}
+	}()
+
+	deadline, cancelFunc := context.WithDeadline(context.Background(), time.Now().Add(5*time.Second))
+	defer cancelFunc()
+	for {
+		select {
+		case <-deadline.Done():
+			break
+		case <-time.After(50 * time.Millisecond):
+			db, err := sql.Open("postgres", fmt.Sprintf("host=localhost port=%d database=pgcat user=admin_user password=admin_pass sslmode=disable", port))
+			if err != nil {
+				continue
+			}
+			rows, err := db.QueryContext(deadline, "SHOW STATS")
+			if err != nil {
+				continue
+			}
+			_ = rows.Close()
+			_ = db.Close()
+			break
+		}
+		break
+	}
+
+	return func() {
+		err := cmd.Process.Signal(os.Interrupt)
+		if err != nil {
+			t.Fatalf("could not interrupt pgcat: %+v", err)
+		}
+		err = os.Remove(cfg.Name())
+		if err != nil {
+			t.Fatalf("could not remove temp file: %+v", err)
+		}
+	}
+}

tests/pgbench/simple.sql

@@ -36,4 +36,4 @@ SELECT abalance FROM pgbench_accounts WHERE aid = :aid;
 SET SERVER ROLE TO 'replica';
 
 -- Read load balancing
-SELECT abalance FROM pgbench_accounts WHERE aid = :aid;
+SELECT abalance FROM pgbench_accounts WHERE aid = :aid;

tests/python/requirements.txt

@@ -1,2 +1,3 @@
+pytest
 psycopg2==2.9.3
-psutil==5.9.1
+psutil==5.9.1

tests/python/test_auth.py

@@ -0,0 +1,71 @@
+import utils
+import signal
+
+class TestTrustAuth:
+    @classmethod
+    def setup_method(cls):
+        config= """
+        [general]
+        host = "0.0.0.0"
+        port = 6432
+        admin_username = "admin_user"
+        admin_password = ""
+        admin_auth_type = "trust"
+
+        [pools.sharded_db.users.0]
+        username = "sharding_user"
+        password = "sharding_user"
+        auth_type = "trust"
+        pool_size = 10
+        min_pool_size = 1
+        pool_mode = "transaction"
+
+        [pools.sharded_db.shards.0]
+        servers = [
+          [ "127.0.0.1", 5432, "primary" ],
+        ]
+        database = "shard0"
+        """
+        utils.pgcat_generic_start(config)
+
+    @classmethod
+    def teardown_method(self):
+        utils.pg_cat_send_signal(signal.SIGTERM)
+
+    def test_admin_trust_auth(self):
+        conn, cur = utils.connect_db_trust(admin=True)
+        cur.execute("SHOW POOLS")
+        res = cur.fetchall()
+        print(res)
+        utils.cleanup_conn(conn, cur)
+
+    def test_normal_trust_auth(self):
+        conn, cur = utils.connect_db_trust(autocommit=False)
+        cur.execute("SELECT 1")
+        res = cur.fetchall()
+        print(res)
+        utils.cleanup_conn(conn, cur)
+
+class TestMD5Auth:
+    @classmethod
+    def setup_method(cls):
+        utils.pgcat_start()
+
+    @classmethod
+    def teardown_method(self):
+        utils.pg_cat_send_signal(signal.SIGTERM)
+
+    def test_normal_db_access(self):
+        conn, cur = utils.connect_db(autocommit=False)
+        cur.execute("SELECT 1")
+        res = cur.fetchall()
+        print(res)
+        utils.cleanup_conn(conn, cur)
+
+    def test_admin_db_access(self):
+        conn, cur = utils.connect_db(admin=True)
+
+        cur.execute("SHOW POOLS")
+        res = cur.fetchall()
+        print(res)
+        utils.cleanup_conn(conn, cur)

tests/python/test_pgcat.py

@@ -1,84 +1,12 @@
-from typing import Tuple
-import psycopg2
-import psutil
-import os
+
 import signal
 import time
 
+import psycopg2
+import utils
+
 SHUTDOWN_TIMEOUT = 5
 
-PGCAT_HOST = "127.0.0.1"
-PGCAT_PORT = "6432"
-
-
-def pgcat_start():
-    pg_cat_send_signal(signal.SIGTERM)
-    os.system("./target/debug/pgcat .circleci/pgcat.toml &")
-    time.sleep(2)
-
-
-def pg_cat_send_signal(signal: signal.Signals):
-    try:
-        for proc in psutil.process_iter(["pid", "name"]):
-            if "pgcat" == proc.name():
-                os.kill(proc.pid, signal)
-    except Exception as e:
-        # The process can be gone when we send this signal
-        print(e)
-
-    if signal == signal.SIGTERM:
-        # Returns 0 if pgcat process exists
-        time.sleep(2)
-        if not os.system('pgrep pgcat'):
-            raise Exception("pgcat not closed after SIGTERM")
-
-
-def connect_db(
-    autocommit: bool = True,
-    admin: bool = False,
-) -> Tuple[psycopg2.extensions.connection, psycopg2.extensions.cursor]:
-
-    if admin:
-        user = "admin_user"
-        password = "admin_pass"
-        db = "pgcat"
-    else:
-        user = "sharding_user"
-        password = "sharding_user"
-        db = "sharded_db"
-
-    conn = psycopg2.connect(
-        f"postgres://{user}:{password}@{PGCAT_HOST}:{PGCAT_PORT}/{db}?application_name=testing_pgcat",
-        connect_timeout=2,
-    )
-    conn.autocommit = autocommit
-    cur = conn.cursor()
-
-    return (conn, cur)
-
-
-def cleanup_conn(conn: psycopg2.extensions.connection, cur: psycopg2.extensions.cursor):
-    cur.close()
-    conn.close()
-
-
-def test_normal_db_access():
-    pgcat_start()
-    conn, cur = connect_db(autocommit=False)
-    cur.execute("SELECT 1")
-    res = cur.fetchall()
-    print(res)
-    cleanup_conn(conn, cur)
-
-
-def test_admin_db_access():
-    conn, cur = connect_db(admin=True)
-
-    cur.execute("SHOW POOLS")
-    res = cur.fetchall()
-    print(res)
-    cleanup_conn(conn, cur)
-
 
 def test_shutdown_logic():
 
@@ -86,17 +14,17 @@ def test_shutdown_logic():
     # NO ACTIVE QUERIES SIGINT HANDLING
 
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and send query (not in transaction)
-    conn, cur = connect_db()
+    conn, cur = utils.connect_db()
 
     cur.execute("BEGIN;")
     cur.execute("SELECT 1;")
     cur.execute("COMMIT;")
 
     # Send sigint to pgcat
-    pg_cat_send_signal(signal.SIGINT)
+    utils.pg_cat_send_signal(signal.SIGINT)
     time.sleep(1)
 
     # Check that any new queries fail after sigint since server should close with no active transactions
@@ -108,18 +36,18 @@ def test_shutdown_logic():
         # Fail if query execution succeeded
         raise Exception("Server not closed after sigint")
 
-    cleanup_conn(conn, cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(conn, cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # NO ACTIVE QUERIES ADMIN SHUTDOWN COMMAND
 
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction
-    conn, cur = connect_db()
-    admin_conn, admin_cur = connect_db(admin=True)
+    conn, cur = utils.connect_db()
+    admin_conn, admin_cur = utils.connect_db(admin=True)
 
     cur.execute("BEGIN;")
     cur.execute("SELECT 1;")
@@ -138,24 +66,24 @@ def test_shutdown_logic():
         # Fail if query execution succeeded
         raise Exception("Server not closed after sigint")
 
-    cleanup_conn(conn, cur)
-    cleanup_conn(admin_conn, admin_cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(conn, cur)
+    utils.cleanup_conn(admin_conn, admin_cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # HANDLE TRANSACTION WITH SIGINT
 
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction
-    conn, cur = connect_db()
+    conn, cur = utils.connect_db()
 
     cur.execute("BEGIN;")
     cur.execute("SELECT 1;")
 
     # Send sigint to pgcat while still in transaction
-    pg_cat_send_signal(signal.SIGINT)
+    utils.pg_cat_send_signal(signal.SIGINT)
     time.sleep(1)
 
     # Check that any new queries succeed after sigint since server should still allow transaction to complete
@@ -165,18 +93,18 @@ def test_shutdown_logic():
         # Fail if query fails since server closed
         raise Exception("Server closed while in transaction", e.pgerror)
 
-    cleanup_conn(conn, cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(conn, cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # HANDLE TRANSACTION WITH ADMIN SHUTDOWN COMMAND
 
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction
-    conn, cur = connect_db()
-    admin_conn, admin_cur = connect_db(admin=True)
+    conn, cur = utils.connect_db()
+    admin_conn, admin_cur = utils.connect_db(admin=True)
 
     cur.execute("BEGIN;")
     cur.execute("SELECT 1;")
@@ -194,30 +122,30 @@ def test_shutdown_logic():
         # Fail if query fails since server closed
         raise Exception("Server closed while in transaction", e.pgerror)
 
-    cleanup_conn(conn, cur)
-    cleanup_conn(admin_conn, admin_cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(conn, cur)
+    utils.cleanup_conn(admin_conn, admin_cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # NO NEW NON-ADMIN CONNECTIONS DURING SHUTDOWN
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction
-    transaction_conn, transaction_cur = connect_db()
+    transaction_conn, transaction_cur = utils.connect_db()
 
     transaction_cur.execute("BEGIN;")
     transaction_cur.execute("SELECT 1;")
 
     # Send sigint to pgcat while still in transaction
-    pg_cat_send_signal(signal.SIGINT)
+    utils.pg_cat_send_signal(signal.SIGINT)
     time.sleep(1)
 
     start = time.perf_counter()
     try:
-        conn, cur = connect_db()
+        conn, cur = utils.connect_db()
         cur.execute("SELECT 1;")
-        cleanup_conn(conn, cur)
+        utils.cleanup_conn(conn, cur)
     except psycopg2.OperationalError as e:
         time_taken = time.perf_counter() - start
         if time_taken > 0.1:
@@ -227,49 +155,49 @@ def test_shutdown_logic():
     else:
         raise Exception("Able connect to database during shutdown")
 
-    cleanup_conn(transaction_conn, transaction_cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(transaction_conn, transaction_cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # ALLOW NEW ADMIN CONNECTIONS DURING SHUTDOWN
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction
-    transaction_conn, transaction_cur = connect_db()
+    transaction_conn, transaction_cur = utils.connect_db()
 
     transaction_cur.execute("BEGIN;")
     transaction_cur.execute("SELECT 1;")
 
     # Send sigint to pgcat while still in transaction
-    pg_cat_send_signal(signal.SIGINT)
+    utils.pg_cat_send_signal(signal.SIGINT)
     time.sleep(1)
 
     try:
-        conn, cur = connect_db(admin=True)
+        conn, cur = utils.connect_db(admin=True)
         cur.execute("SHOW DATABASES;")
-        cleanup_conn(conn, cur)
+        utils.cleanup_conn(conn, cur)
     except psycopg2.OperationalError as e:
         raise Exception(e)
 
-    cleanup_conn(transaction_conn, transaction_cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(transaction_conn, transaction_cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # ADMIN CONNECTIONS CONTINUING TO WORK AFTER SHUTDOWN
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction
-    transaction_conn, transaction_cur = connect_db()
+    transaction_conn, transaction_cur = utils.connect_db()
     transaction_cur.execute("BEGIN;")
     transaction_cur.execute("SELECT 1;")
 
-    admin_conn, admin_cur = connect_db(admin=True)
+    admin_conn, admin_cur = utils.connect_db(admin=True)
     admin_cur.execute("SHOW DATABASES;")
 
     # Send sigint to pgcat while still in transaction
-    pg_cat_send_signal(signal.SIGINT)
+    utils.pg_cat_send_signal(signal.SIGINT)
     time.sleep(1)
 
     try:
@@ -277,24 +205,24 @@ def test_shutdown_logic():
     except psycopg2.OperationalError as e:
         raise Exception("Could not execute admin command:", e)
 
-    cleanup_conn(transaction_conn, transaction_cur)
-    cleanup_conn(admin_conn, admin_cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(transaction_conn, transaction_cur)
+    utils.cleanup_conn(admin_conn, admin_cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
     # HANDLE SHUTDOWN TIMEOUT WITH SIGINT
 
     # Start pgcat
-    pgcat_start()
+    utils.pgcat_start()
 
     # Create client connection and begin transaction, which should prevent server shutdown unless shutdown timeout is reached
-    conn, cur = connect_db()
+    conn, cur = utils.connect_db()
 
     cur.execute("BEGIN;")
     cur.execute("SELECT 1;")
 
     # Send sigint to pgcat while still in transaction
-    pg_cat_send_signal(signal.SIGINT)
+    utils.pg_cat_send_signal(signal.SIGINT)
 
     # pgcat shutdown timeout is set to SHUTDOWN_TIMEOUT seconds, so we sleep for SHUTDOWN_TIMEOUT + 1 seconds
     time.sleep(SHUTDOWN_TIMEOUT + 1)
@@ -308,12 +236,7 @@ def test_shutdown_logic():
         # Fail if query execution succeeded
         raise Exception("Server not closed after sigint and expected timeout")
 
-    cleanup_conn(conn, cur)
-    pg_cat_send_signal(signal.SIGTERM)
+    utils.cleanup_conn(conn, cur)
+    utils.pg_cat_send_signal(signal.SIGTERM)
 
     # - - - - - - - - - - - - - - - - - -
-
-
-test_normal_db_access()
-test_admin_db_access()
-test_shutdown_logic()

tests/python/utils.py

@@ -0,0 +1,110 @@
+import os
+import signal
+import time
+from typing import Tuple
+import tempfile
+
+import psutil
+import psycopg2
+
+PGCAT_HOST = "127.0.0.1"
+PGCAT_PORT = "6432"
+
+
+def _pgcat_start(config_path: str):
+    pg_cat_send_signal(signal.SIGTERM)
+    os.system(f"./target/debug/pgcat {config_path} &")
+    time.sleep(2)
+
+
+def pgcat_start():
+    _pgcat_start(config_path='.circleci/pgcat.toml')
+
+
+def pgcat_generic_start(config: str):
+    tmp = tempfile.NamedTemporaryFile()
+    with open(tmp.name, 'w') as f:
+        f.write(config)
+    _pgcat_start(config_path=tmp.name)
+
+
+def glauth_send_signal(signal: signal.Signals):
+    try:
+        for proc in psutil.process_iter(["pid", "name"]):
+            if proc.name() == "glauth":
+                os.kill(proc.pid, signal)
+    except Exception as e:
+        # The process can be gone when we send this signal
+        print(e)
+
+    if signal == signal.SIGTERM:
+        # Returns 0 if pgcat process exists
+        time.sleep(2)
+        if not os.system('pgrep glauth'):
+            raise Exception("glauth not closed after SIGTERM")
+
+
+def pg_cat_send_signal(signal: signal.Signals):
+    try:
+        for proc in psutil.process_iter(["pid", "name"]):
+            if "pgcat" == proc.name():
+                os.kill(proc.pid, signal)
+    except Exception as e:
+        # The process can be gone when we send this signal
+        print(e)
+
+    if signal == signal.SIGTERM:
+        # Returns 0 if pgcat process exists
+        time.sleep(2)
+        if not os.system('pgrep pgcat'):
+            raise Exception("pgcat not closed after SIGTERM")
+
+
+def connect_db(
+    autocommit: bool = True,
+    admin: bool = False,
+) -> Tuple[psycopg2.extensions.connection, psycopg2.extensions.cursor]:
+
+    if admin:
+        user = "admin_user"
+        password = "admin_pass"
+        db = "pgcat"
+    else:
+        user = "sharding_user"
+        password = "sharding_user"
+        db = "sharded_db"
+
+    conn = psycopg2.connect(
+        f"postgres://{user}:{password}@{PGCAT_HOST}:{PGCAT_PORT}/{db}?application_name=testing_pgcat",
+        connect_timeout=2,
+    )
+    conn.autocommit = autocommit
+    cur = conn.cursor()
+
+    return (conn, cur)
+
+def connect_db_trust(
+    autocommit: bool = True,
+    admin: bool = False,
+) -> Tuple[psycopg2.extensions.connection, psycopg2.extensions.cursor]:
+
+    if admin:
+        user = "admin_user"
+        db = "pgcat"
+    else:
+        user = "sharding_user"
+        db = "sharded_db"
+
+    conn = psycopg2.connect(
+        f"postgres://{user}@{PGCAT_HOST}:{PGCAT_PORT}/{db}?application_name=testing_pgcat",
+        connect_timeout=2,
+    )
+    conn.autocommit = autocommit
+    cur = conn.cursor()
+
+    return (conn, cur)
+
+
+def cleanup_conn(conn: psycopg2.extensions.connection, cur: psycopg2.extensions.cursor):
+    cur.close()
+    conn.close()

tests/ruby/Gemfile.lock

@@ -1,22 +1,33 @@
 GEM
   remote: https://rubygems.org/
   specs:
-    activemodel (7.0.4.1)
-      activesupport (= 7.0.4.1)
-    activerecord (7.0.4.1)
-      activemodel (= 7.0.4.1)
-      activesupport (= 7.0.4.1)
-    activesupport (7.0.4.1)
+    activemodel (7.1.4)
+      activesupport (= 7.1.4)
+    activerecord (7.1.4)
+      activemodel (= 7.1.4)
+      activesupport (= 7.1.4)
+      timeout (>= 0.4.0)
+    activesupport (7.1.4)
+      base64
+      bigdecimal
       concurrent-ruby (~> 1.0, >= 1.0.2)
+      connection_pool (>= 2.2.5)
+      drb
       i18n (>= 1.6, < 2)
       minitest (>= 5.1)
+      mutex_m
       tzinfo (~> 2.0)
     ast (2.4.2)
-    concurrent-ruby (1.1.10)
+    base64 (0.2.0)
+    bigdecimal (3.1.8)
+    concurrent-ruby (1.3.4)
+    connection_pool (2.4.1)
     diff-lcs (1.5.0)
-    i18n (1.12.0)
+    drb (2.2.1)
+    i18n (1.14.5)
       concurrent-ruby (~> 1.0)
-    minitest (5.17.0)
+    minitest (5.25.1)
+    mutex_m (0.2.0)
     parallel (1.22.1)
     parser (3.1.2.0)
       ast (~> 2.4.1)
@@ -24,7 +35,8 @@ GEM
     pg (1.3.2)
     rainbow (3.1.1)
     regexp_parser (2.3.1)
-    rexml (3.2.5)
+    rexml (3.3.6)
+      strscan
     rspec (3.11.0)
       rspec-core (~> 3.11.0)
       rspec-expectations (~> 3.11.0)
@@ -50,10 +62,12 @@ GEM
     rubocop-ast (1.17.0)
       parser (>= 3.1.1.0)
     ruby-progressbar (1.11.0)
+    strscan (3.1.0)
+    timeout (0.4.1)
     toml (0.3.0)
       parslet (>= 1.8.0, < 3.0.0)
     toxiproxy (2.0.1)
-    tzinfo (2.0.5)
+    tzinfo (2.0.6)
       concurrent-ruby (~> 1.0)
     unicode-display_width (2.1.0)
 

tests/ruby/admin_spec.rb

@@ -91,6 +91,27 @@ describe "Admin" do
     end
   end
 
+  [
+    "SHOW ME THE MONEY", 
+    "SHOW ME THE WAY", 
+    "SHOW UP", 
+    "SHOWTIME",
+    "HAMMER TIME",
+    "SHOWN TO BE TRUE",
+    "SHOW ",
+    "SHOW     ",
+    "SHOW 1",
+    ";;;;;"
+  ].each do |cmd|
+    describe "Bad command #{cmd}" do
+      it "does not panic and responds with PG::SystemError" do
+        admin_conn = PG::connect(processes.pgcat.admin_connection_string)
+        expect { admin_conn.async_exec(cmd) }.to raise_error(PG::SystemError).with_message(/Unsupported/)
+        admin_conn.close
+      end
+    end
+  end
+
   describe "PAUSE" do
     it "pauses all pools" do
       admin_conn = PG::connect(processes.pgcat.admin_connection_string)

tests/ruby/load_balancing_spec.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
-require_relative 'spec_helper'
+require_relative "spec_helper"
 
 describe "Random Load Balancing" do
   let(:processes) { Helpers::Pgcat.single_shard_setup("sharded_db", 5) }
@@ -8,7 +8,7 @@ describe "Random Load Balancing" do
     processes.pgcat.shutdown
   end
 
-  context "under regular circumstances" do
+  context("under regular circumstances") do
     it "balances query volume between all instances" do
       conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
 
@@ -22,14 +22,14 @@ describe "Random Load Balancing" do
         failed_count += 1
       end
 
-      expect(failed_count).to eq(0)
+      expect(failed_count).to(eq(0))
       processes.all_databases.map(&:count_select_1_plus_2).each do |instance_share|
-        expect(instance_share).to be_within(expected_share * MARGIN_OF_ERROR).of(expected_share)
+        expect(instance_share).to(be_within(expected_share * MARGIN_OF_ERROR).of(expected_share))
       end
     end
   end
 
-  context "when some replicas are down" do
+  context("when some replicas are down") do
     it "balances query volume between working instances" do
       conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
       expected_share = QUERY_COUNT / (processes.all_databases.count - 2)
@@ -49,9 +49,9 @@ describe "Random Load Balancing" do
       processes.all_databases.each do |instance|
         queries_routed = instance.count_select_1_plus_2
         if processes.replicas[0..1].include?(instance)
-          expect(queries_routed).to eq(0)
+          expect(queries_routed).to(eq(0))
         else
-          expect(queries_routed).to be_within(expected_share * MARGIN_OF_ERROR).of(expected_share)
+          expect(queries_routed).to(be_within(expected_share * MARGIN_OF_ERROR).of(expected_share))
         end
       end
     end
@@ -65,7 +65,7 @@ describe "Least Outstanding Queries Load Balancing" do
     processes.pgcat.shutdown
   end
 
-  context "under homogeneous load" do
+  context("under homogeneous load") do
     it "balances query volume between all instances" do
       conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
 
@@ -79,15 +79,15 @@ describe "Least Outstanding Queries Load Balancing" do
         failed_count += 1
       end
 
-      expect(failed_count).to eq(0)
+      expect(failed_count).to(eq(0))
       processes.all_databases.map(&:count_select_1_plus_2).each do |instance_share|
-        expect(instance_share).to be_within(expected_share * MARGIN_OF_ERROR).of(expected_share)
+        expect(instance_share).to(be_within(expected_share * MARGIN_OF_ERROR).of(expected_share))
       end
     end
   end
 
-  context "under heterogeneous load" do
-    xit "balances query volume between all instances based on how busy they are" do
+  context("under heterogeneous load") do
+    xit("balances query volume between all instances based on how busy they are") do
       slow_query_count = 2
       threads = Array.new(slow_query_count) do
         Thread.new do
@@ -108,31 +108,32 @@ describe "Least Outstanding Queries Load Balancing" do
         failed_count += 1
       end
 
-      expect(failed_count).to eq(0)
+      expect(failed_count).to(eq(0))
       # Under LOQ, we expect replicas running the slow pg_sleep
       # to get no selects
       expect(
-        processes.
-          all_databases.
-          map(&:count_select_1_plus_2).
-          count { |instance_share| instance_share == 0 }
-      ).to eq(slow_query_count)
+        processes
+          .all_databases
+          .map(&:count_select_1_plus_2)
+          .count { |instance_share| instance_share == 0 }
+      )
+        .to(eq(slow_query_count))
 
       # We also expect the quick queries to be spread across
       # the idle servers only
-      processes.
-        all_databases.
-        map(&:count_select_1_plus_2).
-        reject { |instance_share| instance_share == 0 }.
-        each do |instance_share|
-          expect(instance_share).to be_within(expected_share * MARGIN_OF_ERROR).of(expected_share)
-      end
+      processes
+        .all_databases
+        .map(&:count_select_1_plus_2)
+        .reject { |instance_share| instance_share == 0 }
+        .each do |instance_share|
+          expect(instance_share).to(be_within(expected_share * MARGIN_OF_ERROR).of(expected_share))
+        end
 
       threads.map(&:join)
     end
   end
 
-  context "when some replicas are down" do
+  context("when some replicas are down") do
     it "balances query volume between working instances" do
       conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
       expected_share = QUERY_COUNT / (processes.all_databases.count - 2)
@@ -149,16 +150,106 @@ describe "Least Outstanding Queries Load Balancing" do
         end
       end
 
-      expect(failed_count).to be <= 2
+      expect(failed_count).to(be <= 2)
       processes.all_databases.each do |instance|
         queries_routed = instance.count_select_1_plus_2
         if processes.replicas[0..1].include?(instance)
-          expect(queries_routed).to eq(0)
+          expect(queries_routed).to(eq(0))
         else
-          expect(queries_routed).to be_within(expected_share * MARGIN_OF_ERROR).of(expected_share)
+          expect(queries_routed).to(be_within(expected_share * MARGIN_OF_ERROR).of(expected_share))
         end
       end
     end
   end
 end
 
+describe "Candidate filtering based on `default_pool`" do
+  let(:processes) {
+    Helpers::Pgcat.single_shard_setup("sharded_db", 5, "transaction", "random", "debug", pool_settings)
+  }
+
+  after do
+    processes.all_databases.map(&:reset)
+    processes.pgcat.shutdown
+  end
+
+  context("with default_pool set to replicas") do
+    context("when all replicas are down ") do
+      let(:pool_settings) do
+        {
+          "default_role" => "replica",
+          "replica_to_primary_failover_enabled" => replica_to_primary_failover_enabled
+        }
+      end
+
+      context("with `replica_to_primary_failover_enabled` set to false`") do
+        let(:replica_to_primary_failover_enabled) { false }
+
+        it(
+          "unbans them automatically to prevent false positives in health checks that could make all replicas unavailable"
+        ) do
+          conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
+          failed_count = 0
+          number_of_replicas = processes[:replicas].length
+
+          # Take down all replicas
+          processes[:replicas].each(&:take_down)
+
+          (number_of_replicas + 1).times do |n|
+            conn.async_exec("SELECT 1 + 2")
+          rescue
+            conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
+            failed_count += 1
+          end
+
+          expect(failed_count).to(eq(number_of_replicas + 1))
+          failed_count = 0
+
+          # Ban_time is configured to 60 so this reset will only work
+          # if the replicas are unbanned automatically
+          processes[:replicas].each(&:reset)
+
+          number_of_replicas.times do
+            conn.async_exec("SELECT 1 + 2")
+          rescue
+            conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
+            failed_count += 1
+          end
+
+          expect(failed_count).to(eq(0))
+        end
+      end
+
+      context("with `replica_to_primary_failover_enabled` set to true`") do
+        let(:replica_to_primary_failover_enabled) { true }
+
+        it "does not unbans them automatically" do
+          conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
+          failed_count = 0
+          number_of_replicas = processes[:replicas].length
+
+          # We need to allow pgcat to open connections to replicas
+          (number_of_replicas + 10).times do |n|
+            conn.async_exec("SELECT 1 + 2")
+          rescue
+            conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
+            failed_count += 1
+          end
+          expect(failed_count).to(eq(0))
+
+          # Take down all replicas
+          processes[:replicas].each(&:take_down)
+
+          (number_of_replicas + 10).times do |n|
+            conn.async_exec("SELECT 1 + 2")
+          rescue
+            conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
+            failed_count += 1
+          end
+
+          expect(failed_count).to(eq(number_of_replicas))
+        end
+      end
+    end
+  end
+end

tests/ruby/prepared_spec.rb

@@ -1,29 +1,214 @@
 require_relative 'spec_helper'
 
 describe 'Prepared statements' do
-  let(:processes) { Helpers::Pgcat.three_shard_setup('sharded_db', 5) }
+  let(:pool_size) { 5 }
+  let(:processes) { Helpers::Pgcat.single_instance_setup("sharded_db", pool_size) }
+  let(:prepared_statements_cache_size) { 100 }
+  let(:server_round_robin) { false }
 
-  context 'enabled' do
-    it 'will work over the same connection' do
+  before do
+    new_configs = processes.pgcat.current_config
+    new_configs["general"]["server_round_robin"] = server_round_robin
+    new_configs["pools"]["sharded_db"]["prepared_statements_cache_size"] = prepared_statements_cache_size
+    new_configs["pools"]["sharded_db"]["users"]["0"]["pool_size"] =  pool_size
+    processes.pgcat.update_config(new_configs)
+    processes.pgcat.reload_config
+  end
+
+  context 'when trying prepared statements' do
+    it 'it allows unparameterized statements to succeed' do
+      conn1 = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+      conn2 = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+
+      prepared_query = "SELECT 1"
+      
+      # prepare query on server 1 and client 1
+      conn1.prepare('statement1', prepared_query)
+      conn1.exec_prepared('statement1')
+      
+      conn2.transaction do
+        # Claim server 1 with client 2
+        conn2.exec("SELECT 2")
+      
+        # Client 1 now runs the prepared query, and it's automatically
+        # prepared on server 2
+        conn1.prepare('statement2', prepared_query)
+        conn1.exec_prepared('statement2')
+      
+        # Client 2 now prepares the same query that was already
+        # prepared on server 1. And PgBouncer reuses that already
+        # prepared query for this different client.
+        conn2.prepare('statement3', prepared_query)
+        conn2.exec_prepared('statement3')
+        end
+    ensure
+        conn1.close if conn1
+        conn2.close if conn2
+    end
+
+    it 'it allows parameterized statements to succeed' do
+      conn1 = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+      conn2 = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+
+      prepared_query = "SELECT $1"
+      
+      # prepare query on server 1 and client 1
+      conn1.prepare('statement1', prepared_query)
+      conn1.exec_prepared('statement1', [1])
+    
+      conn2.transaction do
+        # Claim server 1 with client 2
+        conn2.exec("SELECT 2")
+      
+        # Client 1 now runs the prepared query, and it's automatically
+        # prepared on server 2
+        conn1.prepare('statement2', prepared_query)
+        conn1.exec_prepared('statement2', [1])
+      
+        # Client 2 now prepares the same query that was already
+        # prepared on server 1. And PgBouncer reuses that already
+        # prepared query for this different client.
+        conn2.prepare('statement3', prepared_query)
+        conn2.exec_prepared('statement3', [1])
+      end
+    ensure
+       conn1.close if conn1
+       conn2.close if conn2
+
+    end
+  end
+
+  context 'when trying large packets' do
+    it "works with large parse" do
+      conn1 = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+
+      long_string = "1" * 4096 * 10
+      prepared_query = "SELECT '#{long_string}'"
+    
+  
+      # prepare query on server 1 and client 1
+      conn1.prepare('statement1', prepared_query)
+      result = conn1.exec_prepared('statement1')
+    
+      # assert result matches long_string
+      expect(result.getvalue(0, 0)).to eq(long_string)
+    ensure
+      conn1.close if conn1
+    end
+
+    it "works with large bind" do
+      conn1 = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+    
+      long_string = "1" * 4096 * 10
+      prepared_query = "SELECT $1::text"
+    
+      # prepare query on server 1 and client 1
+      conn1.prepare('statement1', prepared_query)
+      result = conn1.exec_prepared('statement1', [long_string])
+    
+      # assert result matches long_string
+      expect(result.getvalue(0, 0)).to eq(long_string)
+    ensure
+      conn1.close if conn1
+    end
+  end
+
+  context 'when statement cache is smaller than set of unqiue statements' do
+    let(:prepared_statements_cache_size) { 1 }
+    let(:pool_size) { 1 }
+
+    it "evicts all but 1 statement from the server cache" do
+      conn = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+
+      5.times do |i|
+        prepared_query = "SELECT '#{i}'"
+        conn.prepare("statement#{i}", prepared_query)
+        result = conn.exec_prepared("statement#{i}")
+        expect(result.getvalue(0, 0)).to eq(i.to_s)
+      end
+
+      # Check number of prepared statements (expected: 1)
+      n_statements = conn.exec("SELECT count(*) FROM pg_prepared_statements").getvalue(0, 0).to_i
+      expect(n_statements).to eq(1)
+    end
+  end
+
+  context 'when statement cache is larger than set of unqiue statements' do
+    let(:pool_size) { 1 }
+
+    it "does not evict any of the statements from the cache" do
+      # cache size 5
+      conn = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+
+      5.times do |i|
+        prepared_query = "SELECT '#{i}'"
+        conn.prepare("statement#{i}", prepared_query)
+        result = conn.exec_prepared("statement#{i}")
+        expect(result.getvalue(0, 0)).to eq(i.to_s)
+      end
+
+      # Check number of prepared statements (expected: 1)
+      n_statements = conn.exec("SELECT count(*) FROM pg_prepared_statements").getvalue(0, 0).to_i
+      expect(n_statements).to eq(5)
+    end
+  end
+
+  context 'when preparing the same query' do
+    let(:prepared_statements_cache_size) { 5 }
+    let(:pool_size) { 5 }
+
+    it "reuses statement cache when there are different statement names on the same connection" do
       conn = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
 
       10.times do |i|
         statement_name = "statement_#{i}"
         conn.prepare(statement_name, 'SELECT $1::int')
         conn.exec_prepared(statement_name, [1])
-        conn.describe_prepared(statement_name)
       end
+
+      # Check number of prepared statements (expected: 1)
+      n_statements = conn.exec("SELECT count(*) FROM pg_prepared_statements").getvalue(0, 0).to_i
+      expect(n_statements).to eq(1)
     end
 
-    it 'will work with new connections' do
-      10.times do
+    it "reuses statement cache when there are different statement names on different connections" do
+      10.times do |i|
         conn = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
-
-        statement_name = 'statement1'
-        conn.prepare('statement1', 'SELECT $1::int')
-        conn.exec_prepared('statement1', [1])
-        conn.describe_prepared('statement1')
+        statement_name = "statement_#{i}"
+        conn.prepare(statement_name, 'SELECT $1::int')
+        conn.exec_prepared(statement_name, [1])
       end
+
+      # Check number of prepared statements (expected: 1)
+      conn = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+      n_statements = conn.exec("SELECT count(*) FROM pg_prepared_statements").getvalue(0, 0).to_i
+      expect(n_statements).to eq(1)
+    end
+  end
+
+  context 'when reloading config' do
+    let(:pool_size) { 1 }
+
+    it "test_reload_config" do
+      conn = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+
+      # prepare query
+      conn.prepare('statement1', 'SELECT 1')
+      conn.exec_prepared('statement1')
+
+      # Reload config which triggers pool recreation
+      new_configs = processes.pgcat.current_config
+      new_configs["pools"]["sharded_db"]["prepared_statements_cache_size"] = prepared_statements_cache_size + 1
+      processes.pgcat.update_config(new_configs)
+      processes.pgcat.reload_config
+
+      # check that we're starting with no prepared statements on the server
+      conn_check = PG.connect(processes.pgcat.connection_string('sharded_db', 'sharding_user'))
+      n_statements = conn_check.exec("SELECT count(*) FROM pg_prepared_statements").getvalue(0, 0).to_i
+      expect(n_statements).to eq(0)
+
+      # still able to run prepared query
+      conn.exec_prepared('statement1')
     end
   end
 end

tests/ruby/stats_spec.rb

@@ -233,17 +233,19 @@ describe "Stats" do
         sleep(1.1) # Allow time for stats to update
         admin_conn = PG::connect(processes.pgcat.admin_connection_string)
         results = admin_conn.async_exec("SHOW POOLS")[0]
-        %w[cl_idle cl_cancel_req sv_idle sv_used sv_tested sv_login maxwait].each do |s|
+
+        %w[cl_idle cl_cancel_req sv_idle sv_used sv_tested sv_login].each do |s|
           raise StandardError, "Field #{s} was expected to be 0 but found to be #{results[s]}" if results[s] != "0"
         end
 
+        expect(results["maxwait"]).to eq("1")
         expect(results["cl_waiting"]).to eq("2")
         expect(results["cl_active"]).to eq("2")
         expect(results["sv_active"]).to eq("2")
 
         sleep(2.5) # Allow time for stats to update
         results = admin_conn.async_exec("SHOW POOLS")[0]
-        %w[cl_active cl_waiting cl_cancel_req sv_active sv_used sv_tested sv_login].each do |s|
+        %w[cl_active cl_waiting cl_cancel_req sv_active sv_used sv_tested sv_login maxwait].each do |s|
           raise StandardError, "Field #{s} was expected to be 0 but found to be #{results[s]}" if results[s] != "0"
         end
         expect(results["cl_idle"]).to eq("4")
@@ -255,22 +257,23 @@ describe "Stats" do
 
       it "show correct max_wait" do
         threads = []
+        admin_conn = PG::connect(processes.pgcat.admin_connection_string)
         connections = Array.new(4) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
         connections.each do |c|
           threads << Thread.new { c.async_exec("SELECT pg_sleep(1.5)") rescue nil }
         end
+        sleep(1.1)
+        results = admin_conn.async_exec("SHOW POOLS")[0]
+        # Value is only reported when there are clients waiting
+        expect(results["maxwait"]).to eq("1")
+        expect(results["maxwait_us"].to_i).to be_within(20_000).of(100_000)
 
         sleep(2.5) # Allow time for stats to update
-        admin_conn = PG::connect(processes.pgcat.admin_connection_string)
-        results = admin_conn.async_exec("SHOW POOLS")[0]
-
-        expect(results["maxwait"]).to eq("1")
-        expect(results["maxwait_us"].to_i).to be_within(200_000).of(500_000)
-        connections.map(&:close)
-
-        sleep(4.5) # Allow time for stats to update
         results = admin_conn.async_exec("SHOW POOLS")[0]
+        # no clients are waiting so value is 0
         expect(results["maxwait"]).to eq("0")
+        expect(results["maxwait_us"]).to eq("0")
+        connections.map(&:close)
 
         threads.map(&:join)
       end
@@ -329,6 +332,40 @@ describe "Stats" do
       admin_conn.close
       connections.map(&:close)
     end
+
+    context "when client has waited for a server" do
+      let(:processes) { Helpers::Pgcat.single_instance_setup("sharded_db", 2) }
+
+      it "shows correct maxwait" do
+        threads = []
+        connections = Array.new(3) { |i| PG::connect("#{pgcat_conn_str}?application_name=app#{i}") }
+        connections.each do |c|
+          threads << Thread.new { c.async_exec("SELECT pg_sleep(1.5)") rescue nil }
+        end
+
+        sleep(2.5) # Allow time for stats to update
+        admin_conn = PG::connect(processes.pgcat.admin_connection_string)
+        results = admin_conn.async_exec("SHOW CLIENTS")
+
+        normal_client_results = results.reject { |r| r["database"] == "pgcat" }
+
+        non_waiting_clients = normal_client_results.select { |c| c["maxwait"] == "0" }
+        waiting_clients = normal_client_results.select { |c| c["maxwait"].to_i > 0 }
+
+        expect(non_waiting_clients.count).to eq(2)
+        non_waiting_clients.each do |client|
+          expect(client["maxwait_us"].to_i).to be_between(0, 50_000)
+        end
+
+        expect(waiting_clients.count).to eq(1)
+        waiting_clients.each do |client|
+          expect(client["maxwait_us"].to_i).to be_within(200_000).of(500_000)
+        end
+
+        admin_conn.close
+        connections.map(&:close)
+      end
+    end
   end
 
 

tests/rust/src/main.rs

@@ -15,8 +15,13 @@ async fn test_prepared_statements() {
     for _ in 0..5 {
         let pool = pool.clone();
         let handle = tokio::task::spawn(async move {
-            for _ in 0..1000 {
-                sqlx::query("SELECT 1").fetch_all(&pool).await.unwrap();
+            for i in 0..1000 {
+                match sqlx::query(&format!("SELECT {:?}", i % 5)).fetch_all(&pool).await {
+                    Ok(_) => (),
+                    Err(err) => {
+                        panic!("prepared statement error: {}", err);
+                    }
+                }
             }
         });
 

utilities/deb.sh

@@ -22,7 +22,7 @@ mkdir -p "$deb_dir/etc/systemd/system"
 cp target/release/pgcat "$deb_dir/usr/bin/pgcat"
 chmod +x "$deb_dir/usr/bin/pgcat"
 
-cp pgcat.toml "$deb_dir/etc/pgcat.toml"
+cp pgcat.toml "$deb_dir/etc/pgcat.example.toml"
 cp pgcat.service "$deb_dir/etc/systemd/system/pgcat.service"
 
 (cat control | envsubst) > "$deb_dir/DEBIAN/control"