Velero Backups After Failure

[Parami De Silva]

Hi

We observed during a backup, the velero pod ran out of memory and restarted. This caused backup and all subsequent backups (with the restarted pod). How can we ensure, that subsequent backups are running after a failure of a previous backup ..?. Note that we are scheduling backups using a cron expression. 

An attempt to answer my question is here, but it doesnt explicitly specify whether subsequent backups fail after a failed backp

Thanks

Kind Regards

Parami De Silva

[Tiger Kaovilai]

How can we ensure, that subsequent backups are running after a failure (due to memory limits and pod restart)?

By increasing memory limit requested.

https://velero.io/docs/v1.9/customize-installation/#customize-resource-requests-and-limits

--
You received this message because you are subscribed to the Google Groups "Project Velero" group.
To unsubscribe from this group and stop receiving emails from it, send an email to projectveler...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/projectvelero/3c973bc3-28af-4ab6-be66-b7ced461f353n%40googlegroups.com.

[Parami De Silva]

Thanks for your prompt response. 
I am trying to clarify the following scenario as well:

1. A pod provisioned at 128Mi memory requests
2. During backup OOM error kills the pod and restarts
3. Backup CRD status is Fail

4. Increase pod memory requests to 1Gi
5. Next backup schedule

My question is with:

Will the next backup (step 5) run without any issue..? or will it be marked as failed as backup in step 3 failed. 
I want to confirm whether a backup depends on the status of a previous failed backups as mentioned in this thread
The reason we need to check this is to understand how we can use vertical pod autoscaling with velero. 
Thanks 

Parami