Using the Bitnami redis helm chart version 17.7.5 on Kubernetes version 1.24, we sometimes unexpectedly run into the error: Unhealthy: Readiness probe failed: MISCONF Errors writing to the AOF file: Input/output error MISCONF Errors writing to the AOF file: Input/output error
As far as I can tell, the errors happen nondeterministically.
What could be causing this error? If it's too general, what are ways that I can further debug this issue myself? Happy to provide more information if needed.
I've tried upgrading chart/redis versions, looking at resource consumption around the time of failure, digging through node syslogs, but so far no luck.