Open msbutler opened 2 weeks ago
in the 150 node scale test, we do not replan as eagerly as we could, causing many nodes to hang out without any work. unsure if we should revisit this, given online restore.
Here's our current replanning heuristic in restore: https://github.com/msbutler/cockroach/blob/master/pkg/ccl/backupccl/restore_job.go#L475
Jira issue: CRDB-44330
cc @cockroachdb/disaster-recovery
in the 150 node scale test, we do not replan as eagerly as we could, causing many nodes to hang out without any work. unsure if we should revisit this, given online restore.
Here's our current replanning heuristic in restore: https://github.com/msbutler/cockroach/blob/master/pkg/ccl/backupccl/restore_job.go#L475
Jira issue: CRDB-44330