You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@cassandra.apache.org by Philippe <wa...@gmail.com> on 2011/09/26 21:26:04 UTC
GC for ParNew on 0.8.6
Ever since upgrading to 0.8.6, my nodes' system.log is littered with
GCInspector logs such as these
INFO [ScheduledTasks:1] 2011-09-26 21:23:40,468 GCInspector.java (line 122)
GC for ParNew: 209 ms for 1 collections, 4747932608 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:23:43,788 GCInspector.java (line 122)
GC for ParNew: 218 ms for 1 collections, 4963170032 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:23:55,893 GCInspector.java (line 122)
GC for ParNew: 471 ms for 2 collections, 5564471992 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:23:57,095 GCInspector.java (line 122)
GC for ParNew: 569 ms for 2 collections, 5635053360 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:23:58,096 GCInspector.java (line 122)
GC for ParNew: 234 ms for 1 collections, 6101724672 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:23:59,129 GCInspector.java (line 122)
GC for ParNew: 573 ms for 2 collections, 5933842376 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:24:00,130 GCInspector.java (line 122)
GC for ParNew: 295 ms for 1 collections, 6289652528 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:24:01,331 GCInspector.java (line 122)
GC for ParNew: 490 ms for 2 collections, 6183013200 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:24:02,333 GCInspector.java (line 122)
GC for ParNew: 207 ms for 1 collections, 6350919632 used; max is 16838033408
INFO [ScheduledTasks:1] 2011-09-26 21:24:09,464 GCInspector.java (line 122)
GC for ParNew: 440 ms for 2 collections, 6824630136 used; max is 16838033408
I have changed absolutely nothing and the work load is perhaps even lower
than before upgrading because I have paused most updates to the cluster. Did
a log level change or does this have deeper meaning ?
Thanks
Re: GC for ParNew on 0.8.6
Posted by Richard Low <rl...@acunu.com>.
Hi Philippe,
On Thu, Sep 29, 2011 at 6:47 AM, Philippe <wa...@gmail.com> wrote:
> No it was an upgrade from 0.8.4 or 0.8.5 depending on the nodes.
> No cassandra-env files were changed during the update.
> Any other ideas? The cluster has just been weird ever since running 0.8.6 :
> has anyone else upgraded and not run into this?
What do you mean by the cluster has been weird since the upgrade?
Have you noticed slow-downs? Any other messages in the logs that have
appeared since the upgrade?
Richard.
--
Richard Low
Acunu | http://www.acunu.com | @acunu
Re: GC for ParNew on 0.8.6
Posted by Philippe <wa...@gmail.com>.
No it was an upgrade from 0.8.4 or 0.8.5 depending on the nodes.
No cassandra-env files were changed during the update.
Any other ideas? The cluster has just been weird ever since running 0.8.6 :
has anyone else upgraded and not run into this?
Le 28 sept. 2011 09:32, "Peter Schuller" <pe...@infidyne.com> a
écrit :
>> I have changed absolutely nothing and the work load is perhaps even lower
>> than before upgrading because I have paused most updates to the cluster.
Did
>> a log level change or does this have deeper meaning ?
>
> You don't say which version you upgraded from but I suspect 0.7? JVM
> options like heap size calculations have changed. I am guessing you're
> seeing it in the logs because the young generation was sized
> differently, causing young generation garbage collections to be longer
> than before and thus triggering the logging by the Cassandra
> GCInspector.
>
> --
> / Peter Schuller (@scode on twitter)
Re: GC for ParNew on 0.8.6
Posted by Peter Schuller <pe...@infidyne.com>.
> I have changed absolutely nothing and the work load is perhaps even lower
> than before upgrading because I have paused most updates to the cluster. Did
> a log level change or does this have deeper meaning ?
You don't say which version you upgraded from but I suspect 0.7? JVM
options like heap size calculations have changed. I am guessing you're
seeing it in the logs because the young generation was sized
differently, causing young generation garbage collections to be longer
than before and thus triggering the logging by the Cassandra
GCInspector.
--
/ Peter Schuller (@scode on twitter)