Github user uce commented on a diff in the pull request: https://github.com/apache/flink/pull/2608#discussion_r82976383 --- Diff: flink-runtime/src/main/java/org/apache/flink/runtime/checkpoint/ZooKeeperCompletedCheckpointStore.java --- @@ -172,7 +168,7 @@ public void recover() throws Exception { for (int i = 0; i < numberOfInitialCheckpoints - 1; i++) { try { - removeFromZooKeeperAndDiscardCheckpoint(initialCheckpoints.get(i)); + removeSubsumed(initialCheckpoints.get(i)); --- End diff -- Yes. Even more, I think this is generally dangerous. What if a checkpoint is recovered, but the checkpoint cannot be restored, than we will have lost all others. Since we currently only keep a single one anyways, it is not a problem yet.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. ---