I have an Aurora PostgreSQL instance running with a DB.r6g.2xl size in production and a DB.r6g.large for the UAT environment. Recently, I've noticed a 'High Severity' warning on the UAT environment regarding memory usage. Given that this is a testing environment and not production, should I be concerned about this warning? Or is it acceptable to ignore it for now? The warning suggests tuning queries to use less memory or upgrading to a larger instance, but I'm unsure how to approach this since it's just UAT.
3 Answers
Ideally, your UAT should mirror production closely. If cost is a concern, consider shutting it down during off-hours to save some money. Those high memory alerts shouldn't be ignored – using Performance Insights to profile your queries can help you figure out what needs optimization.
What you're facing likely stems from running a larger instance in production. If your UAT encounters similar loads with reduced memory, you'll get alerts. If your code is optimized, you could either ignore the warnings or lighten the load on UAT, but only if it's affecting your testing.
It's really up to you to gauge the importance of this in your UAT. In my experience, users can push back on UAT if it’s lagging, so I generally keep it performing close to production standards.
Thanks for the insight! I appreciate it. We've been monitoring performance, and since it's a separate environment, we haven't noticed any issues. I think we might stick with the current setup since it saves us money, but just curious – if this were a production issue, would increasing the instance size be the only fix?