[jira] Created: (RVM-812) Increase in failures in small heap perf runs since switch to GenImmix as production

Increase in failures in small heap perf runs since switch to GenImmix as production
-----------------------------------------------------------------------------------

                 Key: RVM-812
                 URL: http://jira.codehaus.org/browse/RVM-812
             Project: RVM
          Issue Type: Bug
            Reporter: David Grove
            Priority: Minor
             Fix For: svn head


Since switching to GenImmix for the performance runs, there has been a marked increase in the number of "3GC" runs that fail due to OOM.  We should either figure out why Immix seems to require slightly more memory to run these tests and fix it or increase the heap size used for the tight heap performance tests.  



--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com
_______________________________________________
Jikesrvm-issues mailing list
Jikesrvm-issues@...
https://lists.sourceforge.net/lists/listinfo/jikesrvm-issues

     [ http://jira.codehaus.org/browse/RVM-812?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Steve Blackburn reassigned RVM-812:
-----------------------------------

    Assignee: Steve Blackburn

> Increase in failures in small heap perf runs since switch to GenImmix as production
> -----------------------------------------------------------------------------------
>
>                 Key: RVM-812
>                 URL: http://jira.codehaus.org/browse/RVM-812
>             Project: RVM
>          Issue Type: Bug
>            Reporter: David Grove
>            Assignee: Steve Blackburn
>            Priority: Minor
>             Fix For: svn head
>
>
> Since switching to GenImmix for the performance runs, there has been a marked increase in the number of "3GC" runs that fail due to OOM.  We should either figure out why Immix seems to require slightly more memory to run these tests and fix it or increase the heap size used for the tight heap performance tests.  

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com
_______________________________________________
Jikesrvm-issues mailing list
Jikesrvm-issues@...
https://lists.sourceforge.net/lists/listinfo/jikesrvm-issues

    [ http://jira.codehaus.org/browse/RVM-812?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=172815#action_172815 ]

Steve Blackburn commented on RVM-812:
-------------------------------------

I'm on to it.   THe problem is with defrag and that it has not been configured properly for the generational setting.   I've ironed out a number of problems already and should be committing some more fixes soon.   The problem right now is that we're too reluctant to trigger a defrag in the generational collector.

> Increase in failures in small heap perf runs since switch to GenImmix as production
> -----------------------------------------------------------------------------------
>
>                 Key: RVM-812
>                 URL: http://jira.codehaus.org/browse/RVM-812
>             Project: RVM
>          Issue Type: Bug
>            Reporter: David Grove
>            Assignee: Steve Blackburn
>            Priority: Minor
>             Fix For: svn head
>
>
> Since switching to GenImmix for the performance runs, there has been a marked increase in the number of "3GC" runs that fail due to OOM.  We should either figure out why Immix seems to require slightly more memory to run these tests and fix it or increase the heap size used for the tight heap performance tests.  

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

------------------------------------------------------------------------------
This SF.net email is sponsored by:
High Quality Requirements in a Collaborative Environment.
Download a free trial of Rational Requirements Composer Now!
http://p.sf.net/sfu/www-ibm-com
_______________________________________________
Jikesrvm-issues mailing list
Jikesrvm-issues@...
https://lists.sourceforge.net/lists/listinfo/jikesrvm-issues

     [ http://jira.codehaus.org/browse/RVM-812?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

David Grove updated RVM-812:
----------------------------

    Affects Version/s: 3.1.0
        Fix Version/s:     (was: svn head)
                       3.1.1

> Increase in failures in small heap perf runs since switch to GenImmix as production
> -----------------------------------------------------------------------------------
>
>                 Key: RVM-812
>                 URL: http://jira.codehaus.org/browse/RVM-812
>             Project: RVM
>          Issue Type: Bug
>    Affects Versions: 3.1.0
>            Reporter: David Grove
>            Assignee: Steve Blackburn
>            Priority: Minor
>             Fix For: 3.1.1
>
>
> Since switching to GenImmix for the performance runs, there has been a marked increase in the number of "3GC" runs that fail due to OOM.  We should either figure out why Immix seems to require slightly more memory to run these tests and fix it or increase the heap size used for the tight heap performance tests.  

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

------------------------------------------------------------------------------
OpenSolaris 2009.06 is a cutting edge operating system for enterprises
looking to deploy the next generation of Solaris that includes the latest
innovations from Sun and the OpenSource community. Download a copy and
enjoy capabilities such as Networking, Storage and Virtualization.
Go to: http://p.sf.net/sfu/opensolaris-get
_______________________________________________
Jikesrvm-issues mailing list
Jikesrvm-issues@...
https://lists.sourceforge.net/lists/listinfo/jikesrvm-issues

    [ http://jira.codehaus.org/browse/RVM-812?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=179735#action_179735 ]

Steve Blackburn commented on RVM-812:
-------------------------------------

I did apply a number of systematic fixes for this problem which were quite effective.

Since then I've noticed we still see quite a few failures.   Daniel and I did some analysis (just looking through cattrack, not investigating the individual errors), and they don't appear to be connected to the move to GenImmix.   a) there are underlying random OOME errors that happened both before and after the move to Immix.   b) the total number of failures went up significantly after the move to biased locking, but at first glance, those errors don't appear to be specifically related to biased locking (so it may just be that it has tickled out some longstanding bugs).

So in summary:  I believe I have fixed the OOME errors related to immix (they were due to problems with defrag, which had not been properly tested in a generational setting).  There are quite a few "random" errors at the moment, but they seem correlated with biased locking (though not necessarily directly due to it).

Finally, Daniel has made a stab at producing a cute little graphic which shows bugs as pixels over time, allowing one to visually correlate various bugs with time.  It is not ready for prime time yet, but perhaps he can share the script. :-)

> Increase in failures in small heap perf runs since switch to GenImmix as production
> -----------------------------------------------------------------------------------
>
>                 Key: RVM-812
>                 URL: http://jira.codehaus.org/browse/RVM-812
>             Project: RVM
>          Issue Type: Bug
>    Affects Versions: 3.1.0
>            Reporter: David Grove
>            Assignee: Steve Blackburn
>            Priority: Minor
>             Fix For: 3.1.1
>
>
> Since switching to GenImmix for the performance runs, there has been a marked increase in the number of "3GC" runs that fail due to OOM.  We should either figure out why Immix seems to require slightly more memory to run these tests and fix it or increase the heap size used for the tight heap performance tests.  

--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://jira.codehaus.org/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing
server and web deployment.
http://p.sf.net/sfu/businessobjects
_______________________________________________
Jikesrvm-issues mailing list
Jikesrvm-issues@...
https://lists.sourceforge.net/lists/listinfo/jikesrvm-issues