zhijiangW commented on a change in pull request #12478:
URL: https://github.com/apache/flink/pull/12478#discussion_r437994969
##########
File path:
flink-streaming-java/src/main/java/org/apache/flink/streaming/runtime/tasks/SubtaskCheckpointCoordinatorImpl.java
##########
@@ -221,9 +232,14 @@ public void checkpointState(
// We generally try to emit the checkpoint barrier as soon as
possible to not affect downstream
// checkpoint alignments
+ if (lastCheckpointId >= metadata.getCheckpointId()) {
Review comment:
Yes, the current complexity somehow comes from coupling abort from
handler with abort from RPC together, since `abortedCheckpointIds` will compare
with `lastCheckpointId(aborted id from handler)` in some processes.
But indeed I can not think of an easy solution right now and also have not
thought of critical problems ATM. So let's further discuss it if encountering
problems future.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
