Github user harshach commented on a diff in the pull request:
https://github.com/apache/storm/pull/1528#discussion_r69152524
--- Diff: storm-core/src/clj/org/apache/storm/daemon/supervisor.clj ---
@@ -425,6 +433,13 @@
". State: " state
", Heartbeat: " (pr-str heartbeat))
(shutdown-worker supervisor id)))
+
+ (doseq [storm-id all-downloaded-storm-ids]
--- End diff --
This looks like you removed the check where we used to keep the jars even
the assignement goes away. We had another issue where if the topology is
rebalanced and the assignment goes away
and immediately comes back to the same supervisor , it tries re-download
the jar and if it happens to be a large jar file than there are chances that
supervisor worker gets killed before it even starts as the jar download takes
longer time.
IMO , i don't see a reason to remove the storm jars on a rebalance or if
the assignement goes away. They should be removed when the topology gets killed.
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
