[
https://issues.jboss.org/browse/JGRP-1425?page=com.atlassian.jira.plugin....
]
Francois Billard updated JGRP-1425:
-----------------------------------
Attachment: FD_ALL_time.jpg
Example of timeoutchecker lag : member loss detected after 31s instead of real loss time :
FD_ALL.timeout + new FD_ALL.timeoutchecker_interval
FD_ALL : add specific interval for timeoutchecker task
------------------------------------------------------
Key: JGRP-1425
URL:
https://issues.jboss.org/browse/JGRP-1425
Project: JGroups
Issue Type: Enhancement
Affects Versions: 3.x
Reporter: Francois Billard
Assignee: Bela Ban
Fix For: 3.1
Attachments: FD_ALL_time.jpg
FD_ALL run 2 task :
- task heartbeat_sender_future to send hearbeat
- task timeout_checker_future to check timestamp of hearbeat received
the thread interval parameter is the same for theses two tasks (FD_ALL.interval)
when you have a long interval, the task timeoutchecker are discovering the loss of a
member too late.
the improvement is to decorrelate these two intervals to let timeoutchecker task stick
the real member loss
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see:
http://www.atlassian.com/software/jira