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.

src/stats/client.rs

@@ -1,4 +1,3 @@
-use super::PoolStats;
 use super::{get_reporter, Reporter};
 use atomic_enum::atomic_enum;
 use std::sync::atomic::*;
@@ -34,12 +33,14 @@ pub struct ClientStats {
     pool_name: String,
     connect_time: Instant,
 
-    pool_stats: Arc<PoolStats>,
     reporter: Reporter,
 
     /// Total time spent waiting for a connection from pool, measures in microseconds
     pub total_wait_time: Arc<AtomicU64>,
 
+    /// Maximum time spent waiting for a connection from pool, measures in microseconds
+    pub max_wait_time: Arc<AtomicU64>,
+
     /// Current state of the client
     pub state: Arc<AtomicClientState>,
 
@@ -61,8 +62,8 @@ impl Default for ClientStats {
             application_name: String::new(),
             username: String::new(),
             pool_name: String::new(),
-            pool_stats: Arc::new(PoolStats::default()),
             total_wait_time: Arc::new(AtomicU64::new(0)),
+            max_wait_time: 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)),
@@ -79,11 +80,9 @@ impl ClientStats {
         username: &str,
         pool_name: &str,
         connect_time: Instant,
-        pool_stats: Arc<PoolStats>,
     ) -> Self {
         Self {
             client_id,
-            pool_stats,
             connect_time,
             application_name: application_name.to_string(),
             username: username.to_string(),
@@ -96,8 +95,6 @@ impl ClientStats {
     /// update metrics on the corresponding pool.
     pub fn disconnect(&self) {
         self.reporter.client_disconnecting(self.client_id);
-        self.pool_stats
-            .client_disconnect(self.state.load(Ordering::Relaxed))
     }
 
     /// Register a client with the stats system. The stats system uses client_id
@@ -105,27 +102,20 @@ impl ClientStats {
     pub fn register(&self, stats: Arc<ClientStats>) {
         self.reporter.client_register(self.client_id, stats);
         self.state.store(ClientState::Idle, Ordering::Relaxed);
-        self.pool_stats.cl_idle.fetch_add(1, Ordering::Relaxed);
     }
 
     /// Reports a client is done querying the server and is no longer assigned a server connection
     pub fn idle(&self) {
-        self.pool_stats
-            .client_idle(self.state.load(Ordering::Relaxed));
         self.state.store(ClientState::Idle, Ordering::Relaxed);
     }
 
     /// Reports a client is waiting for a connection
     pub fn waiting(&self) {
-        self.pool_stats
-            .client_waiting(self.state.load(Ordering::Relaxed));
         self.state.store(ClientState::Waiting, Ordering::Relaxed);
     }
 
     /// Reports a client is done waiting for a connection and is about to query the server.
     pub fn active(&self) {
-        self.pool_stats
-            .client_active(self.state.load(Ordering::Relaxed));
         self.state.store(ClientState::Active, Ordering::Relaxed);
     }
 
@@ -144,6 +134,8 @@ impl ClientStats {
     pub fn checkout_time(&self, microseconds: u64) {
         self.total_wait_time
             .fetch_add(microseconds, Ordering::Relaxed);
+        self.max_wait_time
+            .fetch_max(microseconds, Ordering::Relaxed);
     }
 
     /// Report a query executed by a client against a server