Why does my RDS instance experience random slowdowns even under normal load?

Cloud AWS
#1

There is something strange going on with our AWS RDS (PostgreSQL) instance.

It usually works perfectly, but sometimes queries slow down a lot, even though the CPU, memory, and storage metrics all look fine.

So far, I’ve checked a few things:

– There are no long-running queries in the logs.

– Normal storage IOPS and throughput – AWS hasn’t reported any maintenance or failover events

This slowdown usually only lasts a few hours before things go back to normal.

Has anyone ever had random RDS slowdowns happen when things are normal?

What usually causes this, and how do you usually fix it?

#2

Badly designed SQL queries can slow down your RDS server to a crawl. One tool that PostgreSQL makes available to you is the “EXPLAIN” command, so you can see what your slow queries are actually doing internally.

I’d need to see some SQL to tell you any more than that. But things to watch for are complex joins that are not properly indexed. If EXPLAIN demonstrates that the engine is doing full table scans, this alone can explain your problem – without access to useful indexes, these can involve millions of rows due to the effect of the joins.