Zakelly commented on code in PR #25899:
URL: https://github.com/apache/flink/pull/25899#discussion_r1917880925
##########
flink-runtime/src/main/java/org/apache/flink/runtime/checkpoint/CheckpointProperties.java:
##########
@@ -283,6 +293,39 @@ public String toString() {
false, // Retain on suspension
false);
+ private static final CheckpointProperties FULL_CHECKPOINT_NEVER_RETAINED =
+ new CheckpointProperties(
+ false,
+ CheckpointType.FULL_CHECKPOINT,
+ true,
+ true, // Delete on success
+ true, // Delete on cancellation
+ true, // Delete on failure
+ true, // Delete on suspension
+ false);
+
+ private static final CheckpointProperties
FULL_CHECKPOINT_RETAINED_ON_FAILURE =
+ new CheckpointProperties(
+ false,
+ CheckpointType.FULL_CHECKPOINT,
+ true,
+ true, // Delete on success
+ true, // Delete on cancellation
+ false, // Retain on failure
+ true, // Delete on suspension
+ false);
+
+ private static final CheckpointProperties
FULL_CHECKPOINT_RETAINED_ON_CANCELLATION =
+ new CheckpointProperties(
+ false,
+ CheckpointType.FULL_CHECKPOINT,
+ true,
+ true, // Delete on success
+ false, // Retain on cancellation
+ false, // Retain on failure
+ false, // Retain on suspension
+ false);
+
Review Comment:
Ah, I see.
I read through the `CheckpointCoordinator` and it seems the JM will always
propagate incremental checkpoints periodically (not the ones triggered by Rest
API). But actually the TM will do incremental or full ones by configuration it
reads (create `RocksNativeFullSnapshotStrategy` or
`RocksIncrementalSnapshotStrategy` when state backend build). Meaning that in
UI will always show incremental cps even if we disable that. Am I right? If so,
we should make some change in `CheckpointCoordinator`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: issues-unsubscr...@flink.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
