CrateDB Operator for Kubernetes, and automated backups (CLUSTER_BACKUP_IMAGE?)

CrateDB
1

Hi, we have an on-prem deploy of CrateDB in our Kubernetes cluster, and we’re using the operator to deploy it. We have everything working great, tiered storage, data retention policies, etc… However, in the operator I notice there are provisions for enabling backups and restoration as an automated process… I’ve setup backups to S3, but providing bucket credentials, etc, and then my cluster fails to deploy with the Operator throwing the following in it’s logs:
kopf.objects [INFO ] [cratedb/cratedb] Not deploying backup tools because no backup image is defined.

Digging into this, I found that there is a very sparsely documented environment variable called CLUSTER_BACKUP_IMAGE which according to Configuration — CrateDB Kubernetes Operator 2.46.1.dev3+g8c6e945 documentation should point at a docker container which acts both as a prometheus exporter for backup metrics, as well as handles the backup jobs themselves.

I can’t find any other reference to this, nor any docker containers published anywhere that mention this capability.

Anyone able to shed some light on what is supposed to go into that env var to enable backups?

Thanks in advance for any help.

2

Hi @pmumby_flowcoffee

You’re right - the CLUSTER_BACKUP_IMAGE environment variable is used to define the backup sidecar image that handles scheduled backups and exposes Prometheus metrics for them. However, the backup application is currently not open source, and the image is not published in any public container registry.

At the moment, this functionality is only available as part of CrateDB Cloud and select commercial engagements.