>     [ 
> http://track.sipfoundry.org/browse/XX-6619?page=com.atlassian.
jira.plugin.system.issuetabpanels:comment-> 
tabpanel&focusedCommentId=49843#action_49843 ] 
> 
> K Eccles commented on XX-6619:
> ------------------------------
> 
> Examining Apr 5 soaktest memMon reports (r18469), (%MEM and 
> VSZ) peak as shown:
> master 28.4%    1013784  2-09:40:32
> dist   42.8%     915276  1-03:04:02
> 
> The large jumps in memory use are still seen but the 
> continual growth of proxy and registrar that were seen in the 
> Mar 23(r18351) and Mar 26(r18426) memMon reports seem to have 
> disappeared.  I cannot explain why.
> 
> Between Mar 23 and Mar 26, r18415 fix was added.  This would 
> have stopped lost memory when an event could not be put on a 
> queue, but this did not fix the swapping problem.  The Apr 
> 5(r18469) version has an updated version of stunnel plus 
> various Subscribe/Notify/Register (XX-7389), SAA(XX-8054), 
> sipXbridge and sipXconfig changes.  It does not have any 
> fixes added explicitly to solve the memory use problem.
> 
> I don't see how these other changes would have affected proxy 
> memory usage.  It's possible that one or more of these 
> changes somehow impacted overall system behavior or that 
> there was an issue with the soaktest systems, but I am 
> grasping at straws.
> 
> I don't think there is any more to be done in the current state.

As a last ditch effort, have we tried running the load test without logrotate 
to see if it affects anything?

bob
_______________________________________________
sipx-dev mailing list [email protected]
List Archive: http://list.sipfoundry.org/archive/sipx-dev
Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
sipXecs IP PBX -- http://www.sipfoundry.org/

Reply via email to