Repository: flink
Updated Branches:
refs/heads/master 137e61736 -> 94f1d2b16
[FLINK-7579] [Documentation] More accurate description of yarn.application-attempts.
This closes #4642.
Project: http://git-wip-us.apache.org/repos/asf/flink/repo
Commit: http://git-wip-us.apache.org/repos/asf/flink/commit/94f1d2b1
Tree: http://git-wip-us.apache.org/repos/asf/flink/tree/94f1d2b1
Diff: http://git-wip-us.apache.org/repos/asf/flink/diff/94f1d2b1
Branch: refs/heads/master
Commit: 94f1d2b1672ee8f34db9ae85b94bd916ebc38f3e
Parents: 137e617
Author: Sebastian Klemke <packet@nerdheim.de>
Authored: Tue Sep 5 10:03:43 2017 +0200
Committer: Till Rohrmann <trohrmann@apache.org>
Committed: Sun Oct 22 13:43:35 2017 +0200
----------------------------------------------------------------------
docs/ops/jobmanager_high_availability.md | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/flink/blob/94f1d2b1/docs/ops/jobmanager_high_availability.md
----------------------------------------------------------------------
diff --git a/docs/ops/jobmanager_high_availability.md b/docs/ops/jobmanager_high_availability.md
index e73353b..25701a1 100644
--- a/docs/ops/jobmanager_high_availability.md
+++ b/docs/ops/jobmanager_high_availability.md
@@ -174,7 +174,7 @@ In addition to the HA configuration ([see above](#configuration)), you
have to c
<pre>yarn.application-attempts: 10</pre>
-This means that the application can be restarted 10 times before YARN fails the application.
It's important to note that `yarn.resourcemanager.am.max-attempts` is an upper bound for the
application restarts. Therfore, the number of application attempts set within Flink cannot
exceed the YARN cluster setting with which YARN was started.
+This means that the application can be restarted 9 times for failed attempts before YARN
fails the application (9 retries + 1 initial attempt). Additional restarts can be performed
by YARN if required by YARN operations: Preemption, node hardware failures or reboots, or
NodeManager resyncs. These restarts are not counted against `yarn.application-attempts`, see
<a href="http://johnjianfang.blogspot.de/2015/04/the-number-of-maximum-attempts-of-yarn.html">Jian
Fang's blog post</a>. It's important to note that `yarn.resourcemanager.am.max-attempts`
is an upper bound for the application restarts. Therefore, the number of application attempts
set within Flink cannot exceed the YARN cluster setting with which YARN was started.
#### Container Shutdown Behaviour
|