Backup & Restore concept needed for k8s cluster

[garloff]

As operator I need a way to backup and restore all objects of a k8s cluster, so I can recover from disasters (user errors, failed upgrades, catastrophic infra outage, ...).
As operator I also need (a cluster-specific) backup & restore for the management cluster.

Definition of Done

• A proper backup and restore solution is found
• The solution supports all objects within the k8s cluster.
• Various disaster scenarios, such as user errors, failed upgrades, and catastrophic infrastructure outages can be handled
• Documentation with clear instructions on how to use it exists
• The backup and restore solution has been tested and approved by a third person

[PatrickLaabs]

Hi @garloff,

isn't https://velero.io/ a proper solution for this?

Deployed on a management cluster and on the attached clusters, it will prevent 'almost' a zero-downtime, since you're able to backup and restore (i guess it was almost every) kubernetes object.

Best
Patrick