[
https://issues.apache.org/jira/browse/YARN-1198?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14158289#comment-14158289
]
Jian Fang commented on YARN-1198:
---------------------------------
Craig, thanks for your effort. I have already merged in your YARN-1857 and
YARN-1198 patches.
For blacklisting, I think there are both props and cons on whether different
applications should share the blacklisting information or not. There are valid
cases in both cases. For example, if multiple nodes have difficulties to access
one node, it probably is better to share this information among all nodes
because usually it takes a quite long time to cause sock timeout and exhaust
the retry logic from my own experiences. In this way, the hadoop system can
react faster to a problematic node. Certainly, there are other use cases that
the blacklisting only applies to one application. I am fine with the current
design, but expect Hadoop becomes smarter to handle different scenarios, or at
least provide options for users to customize.
When a node is removed from the cluster because of unhealthy, decommission, or
lost, the blacklisted resources should be updated accordingly. Otherwise, new
issues will come out.
> Capacity Scheduler headroom calculation does not work as expected
> -----------------------------------------------------------------
>
> Key: YARN-1198
> URL: https://issues.apache.org/jira/browse/YARN-1198
> Project: Hadoop YARN
> Issue Type: Bug
> Reporter: Omkar Vinit Joshi
> Assignee: Craig Welch
> Attachments: YARN-1198.1.patch, YARN-1198.10.patch,
> YARN-1198.11-with-1857.patch, YARN-1198.11.patch, YARN-1198.2.patch,
> YARN-1198.3.patch, YARN-1198.4.patch, YARN-1198.5.patch, YARN-1198.6.patch,
> YARN-1198.7.patch, YARN-1198.8.patch, YARN-1198.9.patch
>
>
> Today headroom calculation (for the app) takes place only when
> * New node is added/removed from the cluster
> * New container is getting assigned to the application.
> However there are potentially lot of situations which are not considered for
> this calculation
> * If a container finishes then headroom for that application will change and
> should be notified to the AM accordingly.
> * If a single user has submitted multiple applications (app1 and app2) to the
> same queue then
> ** If app1's container finishes then not only app1's but also app2's AM
> should be notified about the change in headroom.
> ** Similarly if a container is assigned to any applications app1/app2 then
> both AM should be notified about their headroom.
> ** To simplify the whole communication process it is ideal to keep headroom
> per User per LeafQueue so that everyone gets the same picture (apps belonging
> to same user and submitted in same queue).
> * If a new user submits an application to the queue then all applications
> submitted by all users in that queue should be notified of the headroom
> change.
> * Also today headroom is an absolute number ( I think it should be normalized
> but then this is going to be not backward compatible..)
> * Also when admin user refreshes queue headroom has to be updated.
> These all are the potential bugs in headroom calculations
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
