[
https://issues.apache.org/jira/browse/YARN-5944?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15709598#comment-15709598
]
Gour Saha commented on YARN-5944:
---------------------------------
[~billie.rinaldi] do we have to set anything to this config
_*yarn.client.failover-max-attempts*_? Seems like it is a client specific
property but when HA is enabled it can override
_yarn.resourcemanager.connect.max-wait.ms_. So not sure what we should do to it.
Here is the description from
https://hadoop.apache.org/docs/r2.7.1/hadoop-yarn/hadoop-yarn-common/yarn-default.xml
-
{quote}
When HA is enabled, the max number of times FailoverProxyProvider should
attempt failover. When set, this overrides the
yarn.resourcemanager.connect.max-wait.ms. When not set, this is inferred from
yarn.resourcemanager.connect.max-wait.ms.
{quote}
Should we explicitly clear the value of _yarn.client.failover-max-attempts_ in
Slider AM?
> Native services AM should remain up if RM is down
> -------------------------------------------------
>
> Key: YARN-5944
> URL: https://issues.apache.org/jira/browse/YARN-5944
> Project: Hadoop YARN
> Issue Type: Sub-task
> Reporter: Billie Rinaldi
> Assignee: Billie Rinaldi
> Attachments: YARN-5944-yarn-native-services.001.patch
>
>
> If the RM is down, the native services AM will retry connecting the to RM
> until yarn.resourcemanager.connect.max-wait.ms has been exceeded. At that
> point, the AM will stop itself. For a long running service, we would prefer
> the AM to stay up while the RM is down.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
