H team , my be pods are constandly restarting and have been terminating with the following error. along with this i am also facing an issue where it seems one partition for a table is corrupted and i am not able to query from it the query just hangs if i try to query from that partition , version - 3.3.17 , would some one be able to help in this case if its a known issue

looks like disk failure?

hi @Kevin Cai is there something that can be done at our end to recover the data , repair table command did not work since i guess tablets were already in Normal state , also would there be a way to prevent the BEs from restarting

If use multiple replications for the table, you can just decommission this node and remove the bad disk, Starrocks should be able to repair replicas from other healthy ones.

@Kevin Cai currently its only single replication , so to prevent restarts we should replace the disk , even if issue identified is only in one partition so far?

replace the disk to recover the node, but the data is lost for sure.