You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mesos.apache.org by "Benjamin Mahler (JIRA)" <ji...@apache.org> on 2014/05/22 21:21:06 UTC
[jira] [Created] (MESOS-1408) Unnecessary queuing of status update
acknowledgments in the scheduler drive.r
Benjamin Mahler created MESOS-1408:
--------------------------------------
Summary: Unnecessary queuing of status update acknowledgments in the scheduler drive.r
Key: MESOS-1408
URL: https://issues.apache.org/jira/browse/MESOS-1408
Project: Mesos
Issue Type: Bug
Components: c++ api, framework, java api, python api
Affects Versions: 0.19.0
Reporter: Benjamin Mahler
Assignee: Benjamin Mahler
Fix For: 0.19.0
We currently 'defer' in order to acknowledge processed status updates in the scheduler driver.
This is unfortunate because it adds unnecessary delay and unnecessary retries, especially when the scheduler's queue is backed up!
It is sufficient to use 'aborted' to see if we can send a status update. Ideally there would be proper locking on this variable rather than the use of 'volatile', but since we're mostly concerned about the driver aborting on the same thread as the status update was processed, I've left it as is.
This will also be beneficial when we send status updates through the master as is required for MESOS-1389 and the other reconciliation tickets.
--
This message was sent by Atlassian JIRA
(v6.2#6252)