Files
pgcat/tests/ruby/stats_spec.rb

370 lines
15 KiB
Ruby
Raw Normal View History

Refactor Pool Stats to be based off of Server/Client stats (#445) What is wrong Stats reported by SHOW POOLS seem to be leaking. We see lingering cl_idle , cl_waiting, and similarly for sv_idle , sv_active. We confirmed that these are reporting issues not actual lingering clients. This behavior is readily reproducible by running while true; do psql "postgres://sharding_user:sharding_user@localhost:6432/sharded_db" -c "SELECT 1" > /dev/null 2>&1 & done Why it happens I wasn't able to get to figure our the reason for the bug but my best guess is that we have race conditions when updating pool-level stats. So even though individual update operations are atomic, we perform a check then update sequence which is not protected by a guard. https://github.com/postgresml/pgcat/blob/main/src/stats/pool.rs#L174-L179 I am also suspecting that using Relaxed ordering might allow this behavior (I changed all operations to use Ordering::SeqCst but still got lingering clients) How to fix Since SHOW POOLS/SHOW SERVER/SHOW CLIENTS all show the current state of the proxy (as opposed to SHOW STATS which show aggregate values), this PR refactors SHOW POOLS to have it construct the results directly from SHOW SERVER and SHOW CLIENT datasets. This reduces the complexity of stat updates and eliminates the need for having locks when updating pool stats as we only care about updating individual client/server states. This will change the semantics of maxwait, so instead of it holding the maxwait time ever encountered by a client (connected or disconnected), it will only consider connected clients which should be okay given PgCat tends to hold on to client connections more than Pgbouncer.
2023-05-23 08:44:49 -05:00
# frozen_string_literal: true
require 'open3'
require_relative 'spec_helper'
describe "Stats" do
let(:processes) { Helpers::Pgcat.single_instance_setup("sharded_db", 10) }
let(:pgcat_conn_str) { processes.pgcat.connection_string("sharded_db", "sharding_user") }
after do
processes.all_databases.map(&:reset)
processes.pgcat.shutdown
end
describe "SHOW STATS" do
context "clients connect and make one query" do
it "updates *_query_time and *_wait_time" do
connections = Array.new(3) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
connections.each do |c|
Thread.new { c.async_exec("SELECT pg_sleep(0.25)") }
end
sleep(1)
connections.map(&:close)
# wait for averages to be calculated, we shouldn't do this too often
sleep(15.5)
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW STATS")[0]
admin_conn.close
expect(results["total_query_time"].to_i).to be_within(200).of(750)
expect(results["avg_query_time"].to_i).to be_within(50).of(250)
expect(results["total_wait_time"].to_i).to_not eq(0)
expect(results["avg_wait_time"].to_i).to_not eq(0)
end
end
end
describe "SHOW POOLS" do
context "bad credentials" do
it "does not change any stats" do
bad_password_url = URI(pgcat_conn_str)
bad_password_url.password = "wrong"
expect { PG::connect("#{bad_password_url.to_s}?application_name=bad_password") }.to raise_error(PG::ConnectionBad)
sleep(1)
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_idle 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["sv_idle"]).to eq("1")
end
end
context "bad database name" do
it "does not change any stats" do
bad_db_url = URI(pgcat_conn_str)
bad_db_url.path = "/wrong_db"
expect { PG::connect("#{bad_db_url.to_s}?application_name=bad_db") }.to raise_error(PG::ConnectionBad)
sleep(1)
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_idle 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["sv_idle"]).to eq("1")
end
end
context "client connects but issues no queries" do
it "only affects cl_idle stats" do
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
before_test = admin_conn.async_exec("SHOW POOLS")[0]["sv_idle"]
connections = Array.new(20) { PG::connect(pgcat_conn_str) }
sleep(1)
results = admin_conn.async_exec("SHOW POOLS")[0]
%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("20")
expect(results["sv_idle"]).to eq(before_test)
connections.map(&:close)
sleep(1.1)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_active cl_idle 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["sv_idle"]).to eq(before_test)
end
end
context "clients connect and make one query" do
it "only affects cl_idle, sv_idle stats" do
connections = Array.new(5) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
connections.each do |c|
Thread.new { c.async_exec("SELECT pg_sleep(2.5)") }
end
sleep(1.1)
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_idle cl_waiting cl_cancel_req sv_idle 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_active"]).to eq("5")
expect(results["sv_active"]).to eq("5")
sleep(3)
results = admin_conn.async_exec("SHOW POOLS")[0]
%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("5")
expect(results["sv_idle"]).to eq("5")
connections.map(&:close)
sleep(1)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_idle 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["sv_idle"]).to eq("5")
end
end
context "client connects and opens a transaction and closes connection uncleanly" do
it "produces correct statistics" do
connections = Array.new(5) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
connections.each do |c|
Thread.new do
c.async_exec("BEGIN")
c.async_exec("SELECT pg_sleep(0.01)")
c.close
end
end
sleep(1.1)
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_idle 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["sv_idle"]).to eq("5")
end
end
context "client fail to checkout connection from the pool" do
it "counts clients as idle" do
new_configs = processes.pgcat.current_config
new_configs["general"]["connect_timeout"] = 500
new_configs["general"]["ban_time"] = 1
new_configs["general"]["shutdown_timeout"] = 1
new_configs["pools"]["sharded_db"]["users"]["0"]["pool_size"] = 1
processes.pgcat.update_config(new_configs)
processes.pgcat.reload_config
threads = []
connections = Array.new(5) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
connections.each do |c|
threads << Thread.new { c.async_exec("SELECT pg_sleep(1)") rescue PG::SystemError }
end
sleep(2)
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW POOLS")[0]
%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("5")
expect(results["sv_idle"]).to eq("1")
threads.map(&:join)
connections.map(&:close)
end
end
context "clients connects and disconnect normally" do
let(:processes) { Helpers::Pgcat.single_instance_setup("sharded_db", 2) }
it 'shows the same number of clients before and after' do
clients_before = clients_connected_to_pool(processes: processes)
threads = []
connections = Array.new(4) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
connections.each do |c|
threads << Thread.new { c.async_exec("SELECT 1") rescue nil }
end
clients_between = clients_connected_to_pool(processes: processes)
expect(clients_before).not_to eq(clients_between)
connections.each(&:close)
clients_after = clients_connected_to_pool(processes: processes)
expect(clients_before).to eq(clients_after)
end
end
context "clients connects and disconnect abruptly" do
let(:processes) { Helpers::Pgcat.single_instance_setup("sharded_db", 10) }
it 'shows the same number of clients before and after' do
threads = []
connections = Array.new(2) { PG::connect("#{pgcat_conn_str}?application_name=one_query") }
connections.each do |c|
threads << Thread.new { c.async_exec("SELECT 1") }
end
clients_before = clients_connected_to_pool(processes: processes)
random_string = (0...8).map { (65 + rand(26)).chr }.join
connection_string = "#{pgcat_conn_str}?application_name=#{random_string}"
faulty_client = Process.spawn("psql -Atx #{connection_string} >/dev/null")
sleep(1)
# psql starts two processes, we only know the pid of the parent, this
# ensure both are killed
`pkill -9 -f '#{random_string}'`
Process.wait(faulty_client)
clients_after = clients_connected_to_pool(processes: processes)
expect(clients_before).to eq(clients_after)
end
end
context "clients overwhelm server pools" do
let(:processes) { Helpers::Pgcat.single_instance_setup("sharded_db", 2) }
it "cl_waiting is updated to show it" do
threads = []
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)") }
end
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|
raise StandardError, "Field #{s} was expected to be 0 but found to be #{results[s]}" if results[s] != "0"
end
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|
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")
expect(results["sv_idle"]).to eq("2")
threads.map(&:join)
connections.map(&:close)
end
it "show correct max_wait" do
threads = []
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(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]
expect(results["maxwait"]).to eq("0")
threads.map(&:join)
end
end
end
describe "SHOW CLIENTS" do
it "reports correct number and application names" do
conn_str = processes.pgcat.connection_string("sharded_db", "sharding_user")
connections = Array.new(20) { |i| PG::connect("#{conn_str}?application_name=app#{i % 5}") }
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
sleep(1) # Wait for stats to be updated
results = admin_conn.async_exec("SHOW CLIENTS")
expect(results.count).to eq(21) # count admin clients
expect(results.select { |c| c["application_name"] == "app3" || c["application_name"] == "app4" }.count).to eq(8)
expect(results.select { |c| c["database"] == "pgcat" }.count).to eq(1)
connections[0..5].map(&:close)
sleep(1) # Wait for stats to be updated
results = admin_conn.async_exec("SHOW CLIENTS")
expect(results.count).to eq(15)
connections[6..].map(&:close)
sleep(1) # Wait for stats to be updated
expect(admin_conn.async_exec("SHOW CLIENTS").count).to eq(1)
admin_conn.close
end
it "reports correct number of queries and transactions" do
conn_str = processes.pgcat.connection_string("sharded_db", "sharding_user")
connections = Array.new(2) { |i| PG::connect("#{conn_str}?application_name=app#{i}") }
connections.each do |c|
c.async_exec("SELECT 1")
c.async_exec("SELECT 2")
c.async_exec("SELECT 3")
c.async_exec("BEGIN")
c.async_exec("SELECT 4")
c.async_exec("SELECT 5")
c.async_exec("COMMIT")
end
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
sleep(1) # Wait for stats to be updated
results = admin_conn.async_exec("SHOW CLIENTS")
expect(results.count).to eq(3)
normal_client_results = results.reject { |r| r["database"] == "pgcat" }
expect(normal_client_results[0]["transaction_count"]).to eq("4")
expect(normal_client_results[1]["transaction_count"]).to eq("4")
expect(normal_client_results[0]["query_count"]).to eq("7")
expect(normal_client_results[1]["query_count"]).to eq("7")
admin_conn.close
connections.map(&:close)
end
end
describe "Query Storm" do
context "when the proxy receives overwhelmingly large number of short quick queries" do
it "should not have lingering clients or active servers" do
new_configs = processes.pgcat.current_config
new_configs["general"]["connect_timeout"] = 500
new_configs["general"]["ban_time"] = 1
new_configs["general"]["shutdown_timeout"] = 1
new_configs["pools"]["sharded_db"]["users"]["0"]["pool_size"] = 1
processes.pgcat.update_config(new_configs)
processes.pgcat.reload_config
Array.new(40) do
Thread.new do
conn = PG.connect(processes.pgcat.connection_string("sharded_db", "sharding_user"))
conn.async_exec("SELECT pg_sleep(0.1)")
rescue PG::SystemError
ensure
conn.close
end
end.each(&:join)
sleep 1
admin_conn = PG::connect(processes.pgcat.admin_connection_string)
results = admin_conn.async_exec("SHOW POOLS")[0]
%w[cl_idle cl_waiting cl_cancel_req 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
admin_conn.close
end
end
end
end