Github user sachingoel0101 commented on the pull request:
https://github.com/apache/flink/pull/979#issuecomment-127968729
I've added a few more message handlers:
1. We never miss a `RUNNING` state between restarts.
2. There is a timeout for repeatedly getting `CANCELED/.ING` or
`FAILED/ING` messages.
Further, I worked around the `receiveTimeout` bug(?) that a timeout message
might be enqueued even if we just received a message. This is done by putting a
tolerance limit of 0.1 times the `JOB_MANAGER_TIMEOUT` and maintaining the last
ping from the `JobManager`.
@tillrohrmann , could you look this over again? Lemme know if there are
still unhandled cases.
---
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.
---
