[
https://issues.apache.org/jira/browse/TEZ-3719?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16005809#comment-16005809
]
Gopal V commented on TEZ-3719:
------------------------------
[~jeagles]: I think I have made all of those changes and increased my AM->RM
heartbeat down to 5s - most of Vertex locks have dropped off my profiler.
> DAGImpl.computeProgress slows down dispatcher and ipc threads
> -------------------------------------------------------------
>
> Key: TEZ-3719
> URL: https://issues.apache.org/jira/browse/TEZ-3719
> Project: Apache Tez
> Issue Type: Bug
> Reporter: Rajesh Balamohan
> Assignee: Gopal V
> Attachments: dispatcher_holding_lock.png, TEZ-3719.1.patch,
> TEZ-3719.wip.patch, tez_computeProgress_waiting_for_lock.png
>
>
> The Vertex readLock is held while computing progress, which prevents
> scheduling of tasks from happening on that vertex (which needs a writeLock).
> Making the locked section do as little work as possible would speedup the AM.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
