Increasing passive ACU usage on RDS Aurora MySQL Serverless v2 clusters

1

We have noticed a sudden increase in costs generated by our RDS Aurora MySQL Serverless v2 clusters that we cannot understand, and would like some help in resolving this issue.

Around February 18th, we noticed that the capacity metric on our instances was not decreasing its capacity to a minimum at night when no one was using it. Before this date, our clusters would usually maintain a minimum of 0.5 ACU during off-peak hours. However, after February 18th, the capacity stayed the entire night at 1.5 ACU, and later to 1 after we capped the max capacity at night.

We have followed the instructions recommended by some users and turned off Performance Insight, but the problem still persists. We could not find any queries generated by a user in the logs of the night where no one usually uses the system, only the ones executed internally by RDS.

We suspect that the reason for the sudden increase in costs may be due to an internal change that AWS made to the service. If this is the case, we would appreciate any information you could provide us on this change so that we can optimize our resources and reduce costs wherever possible.

We would be grateful if you could investigate this matter and provide us with a prompt response. We rely on your expertise to help us resolve this issue and continue to use AWS services effectively.

Thank you for your attention to this matter.

3 Answers
1

We have got the same problem. Starting between Febraury and March, without any changes to our workload, the CPU consumed by Aurora Serverless v2 increased and has remained stable at 3 times than it was before.

answered a year ago
  • Our problem has persisted even after two months. It seems that the issue sometimes resolves after a system update maintenance, but returns after a few days. We are still trying to figure out what is the cause of this sudden change in pattern. It's interesting to note that a similar problem occurred with our Serverless PostgreSQL, but it resolved itself after a few days and hasn't happened again, still don't know how it happened.

1

We have the same problem. Was this issue resolved since then?

Thank you.

andork
answered 6 months ago
  • The issue persists intermittently, but its frequency has reduced after disabling features such as logging, enhanced monitoring, and performance insight on the cluster. However, the problem has not been fully resolved.

0

Hello there,

Apologies for any concern this may have caused! This does seem a little strange, but please note that AWS will notify users of planned maintenance or service changes/updates that may affect their resources. Feel free to check your email inbox or Personal Health Dashboard for any alerts: http://go.aws/aws-hd.

If you haven't already, I'd recommend setting up logging to monitor & troubleshoot. This doc can help you set up & view your cluster activity: https://docs.aws.amazon.com/AmazonRDS/latest/AuroraUserGuide/aurora-serverless-v2-administration.html#aurora-serverless-v2.logging. Our Accounts & Billing team will also be happy to help if you need a hand; they can take a closer look at your resources & provide you with additional insight: http://go.aws/support-center.

- Roxy M.

AWS
EXPERT
answered a year ago
  • Hi, really appreciate that AWS will and should notify the users but users would still need to continue their service and we cant afford to stop our service and wait until your changes complete. Thus AWS still charged the user with anomaly high costs despite AWS is the one that causing CPU & ACU spiked.

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions