TisonKun commented on a change in pull request #9832: [FLINK-11843] Bind
lifespan of Dispatcher to leader session
URL: https://github.com/apache/flink/pull/9832#discussion_r337546924
##########
File path:
flink-runtime/src/main/java/org/apache/flink/runtime/dispatcher/Dispatcher.java
##########
@@ -197,19 +191,37 @@ public void onStart() throws Exception {
onFatalError(exception);
throw exception;
}
+
+ startRecoveredJobs();
Review comment:
I agree. I am rearranging the design document attached in FLINK-10333 for
emphasizing our proposal leader store based high-availability service, which
includes the removal of `JobGraphListener` since it is originally introduced
for patching coordinating concurrent modification.
I am going to start a discuss thread recently to see if our community is
glad to accept it. Does it sound good to you? If so, what is a good time to
start this thread?(I can finish the preparation in this week) I'm afraid that
the community is blocked by many ongoing threads but IMO
stability(high-availability) is critical component we'd better take more
priority with.
----------------------------------------------------------------
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]
With regards,
Apache Git Services
