FoundationDB / fdb-kubernetes-operator

A kubernetes operator for FoundationDB
Apache License 2.0
245 stars 82 forks source link

Add support for consistency check role #1546

Open manfontan opened 1 year ago

manfontan commented 1 year ago

What would you like to be added/changed?

Docs: https://apple.github.io/foundationdb/command-line-interface.html#consistencycheck This feature will be helpful to troubleshoot corruption issues in the cluster.

manfontan commented 1 year ago

The idea is to add the consistency check role to the supported roles by the operator. So consistency check can be enabled through Manifest instead of manually as described here: https://github.com/apple/foundationdb/issues/7933

brownleej commented 1 year ago

This is a little more complex than it might sound, because ConsistencyCheck is a fundamentally different kind of worker than the other roles. That's why it's configured explicitly with -r instead of hinted with -c and recruited dynamically. That also means it has to be managed differently. For instance, I believe it will not be considered a normal worker of the cluster in the process status, and can't be restarted with fdbcli.

johscheuer commented 1 year ago

Changing this one to documentation. We should provide a Deployment for everyone that wants to run the consistency checker.

We can think about integrating this into the operator but that woulds require some additional changes on how that role is managed.

Rjerk commented 11 months ago

If the consistency_check finds some corruption, what we should do? @johscheuer @manfontan