Repository navigation
IGNITE-29267 Fix flaky SnapshotCompatibilityTest#testSnapshotRestore - #13808
Open
NSAmelchev wants to merge 1 commit into
Open
NSAmelchev wants to merge 1 commit into
NSAmelchev wants to merge 1 commit into
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
SnapshotCompatibilityTest#testSnapshotRestoreis flaky: a snapshot restore started right after the cluster activation fails with "Cache group restore operation was rejected. The cluster should be active."IgniteCluster#statereturns when all nodes have finished the state change exchange, but the local cluster state stays in transition untilChangeGlobalStateFinishMessageis processed.SnapshotRestoreProcess#startchecked the transition flag in the caller's thread and rejected the operation.The transition check is removed from
SnapshotRestoreProcess#start. The cluster state is validated again on the prepare phase, which is ordered with the finish message by discovery, so the restore is still rejected if the state change is not finished by then.Behavior change: in that case the restore is rejected with the same message on the prepare phase, i.e. after the snapshot check, instead of immediately.
The workaround for the same race is removed from
IgniteClusterSnapshotRestoreSelfTest#testUserClassRestored.