Postmortem: how we exhausted the connection pool in llamaindex-tools
It took me two weeks of on-and-off digging and plenty of wrong turns. Writing the process down as it happened so the next person spends less time.
// Minimal reproduction: you must use a real long-tail distribution here.
// Uniform load-test traffic will never trigger this.
func (s *Server) handle(ctx context.Context) error {
conn, err := s.pool.Acquire(ctx)
if err != nil {
return fmt.Errorf("acquire: %w", err)
}
defer conn.Release()
return s.do(ctx, conn)
}The first thing was to collapse the variables. We were changing config and upgrading the version at the same time, and afterwards nobody could say which change caused what. We rolled back to moving one variable at a time, re-ran three times, and only then did the curve settle. Tedious, but not skippable.
What genuinely surprised me was the tail. The average looked great while P99 jumped by an order of magnitude past some threshold. The cause was not llamaindex-tools itself but our upstream connection reuse — the load test traffic was too clean and hid the long-tail requests.
On trade-offs, my view is this: if nobody on the team owns this area long-term, do not introduce a second mechanism. With two coexistence you first have to work out which one is even in play when things break, and that costs far more than the performance you saved.
Worth noting: the official docs do cover this, just in a very inconspicuous spot. I only found it reading the source comments, where the author explains the reasoning — roughly "so that it degrades into predictable behaviour in extreme cases".