RDS Credit Balance suddenly dropped to Zero!
I have an RDS db.t4g.micro RDS Postgres instance that has been running without issue for several months now. Since launch, the Credit Balance accrued to 288 within a day or so, and has maintained this value ever since.
I have a cloudwatch alarm to notify me if the value dips below a certain value, and today I received a notice that my Credit Balance was Zero (0.0)!
Looking at the metrics, I can see that there is a 10-minute gap of missing data, and then the credit balance being reported is 0.0 once reporting has resumed. And now, the reporting is showing the credit balance is accrusing again, but from Zero. I also checked CPU load and other metrics and nothing is seen out of expected values to indicate any sort of spike in resources that could deplete 288 credits in only 10 minutes!
I assume this is some sort of bug?
- Newest
- Most votes
- Most comments
Hmm, seems like it may have been related to an underlying hardware or node failure:
https://repost.aws/questions/QU_aWQmDMvR4qzoLbhgUf6EA/rds-went-into-recovery-by-itself
Relevant content
- asked a year ago
- asked 2 years ago
- asked 5 years ago
AWS OFFICIALUpdated 2 years ago- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated 2 years ago
It looks like my RDS instance experienced some sort of reboot event, resulting in database recovery. And when that happens, I believe all CPU credits are lost/reset. I am not sure what happened, but my RDS event log shows such an event occured in the timeframe that is "missing". Unfortunately I don't see anything in the other metrics that would cause this to occur: