Aurora serverless v2 instance doesn't scale down to minimum 0.5 ACU + random spikes

Hello there,

I understand that you have configured an Aurora Serverless V2 PostgreSQL cluster but it doesn't scale down to minimum ACU set by you i.e 0.5. Further you upgraded the cluster from version 13.10 to 13.11 but the issue still persist.

Your cluster has below instances:-

1) 1 provisioned writer instance which is under constant load.
2) 1 provisioned reader instance which is also under constant load.
3) 1 serverless reader instance which you use only for specific reading queries during peak time to serve the unpredictable and spiky load.

Issue description:- When you don't send requests to the serverless instance it doesn't scale down to 0.5 but stays at around 2.5-3.5 ACUs.

Steps taken by you in order to mitigate the issue:-

-- You checked and you are sure that the serverless v2 instance is not queried by you and all queries are redirected to the writer instance during the off peak hours.

-- We turned off the Performance insight monitoring on the serverless instance.

-- The log export is also turned off on the serverless instance.

-- max_connections and max_locks_per_transaction is set to the default values.

-- Failover priority is set to tier-2 on the serverless instance Failover priority

Looking at the mitigation steps and the screenshots shared by you I understand you have followed the troublehsooting steps mentioned in below documentation but the ACU still didn't scale down to 0.5.

[+] Performance and scaling for Aurora Serverless v2 - Troubleshooting Aurora Serverless v2 capacity issues - https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless-v2.setting-capacity.html#aurora-serverless-v2.troubleshooting

Please allow me explain that, as for the issue that you are facing as the steps mentioned in public documentation are already been taken care of, to further troubleshoot the resources needs to be checked and AWS support engineers have the necessary tools and permissions to dive deep and provide exact cause as to why ACU is not scaling down.

That being said, since this issue might need deeper investigation you can always open a support case with AWS Support Engineering Team. Certain non-public information might be required such as your RDS instance and metric details and information regarding your database usage hence you can open a support case with AWS using the following support link and this issue can be investigated for further troubleshooting.

[+] : AWS Support team - https://console.aws.amazon.com/support/

I sincerely hope above information is helpful for you. Have a great rest of your week!