hi guys, incase of a cluster break down, is there a way to restore and map the existing segments i have deep stored in s3 to the new cluster? any docs available on this?

Answered here https://apache-pinot.slack.com/archives/C01H1S9J5BJ/p1682002649914919

@Mayank @Mark Needham - the steps described here allowed me to successfully restore data from cluster A(old) to cluster B(new) but the requirement here was that cluster A was fully functional and had all the swagger APIs working. is there anyway where I have a broken cluster A and i can still map data of the broken cluster to the new cluster?

What does broken mean? If you have the data aim deepstore and metadata in zk, you should be able to.

@Mayank how to access the zookeeper metadata assuming the controller is broken and i cannot port-forward it to localhost

You can directly access zk port. But the idea there is that you have snapshots enabled in zk which can be used to start the new zk deployment

Interesting. How to test out these snaps? Is there any doc for this? @Mark Needham @Mayank

Yes, Zookeeper docs should definitely have it