Return-Path: Delivered-To: apmail-cassandra-user-archive@www.apache.org Received: (qmail 22150 invoked from network); 20 Apr 2010 17:35:52 -0000 Received: from unknown (HELO mail.apache.org) (140.211.11.3) by 140.211.11.9 with SMTP; 20 Apr 2010 17:35:52 -0000 Received: (qmail 31152 invoked by uid 500); 20 Apr 2010 17:35:51 -0000 Delivered-To: apmail-cassandra-user-archive@cassandra.apache.org Received: (qmail 31036 invoked by uid 500); 20 Apr 2010 17:35:50 -0000 Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: user@cassandra.apache.org Delivered-To: mailing list user@cassandra.apache.org Received: (qmail 31021 invoked by uid 99); 20 Apr 2010 17:35:50 -0000 Received: from athena.apache.org (HELO athena.apache.org) (140.211.11.136) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 20 Apr 2010 17:35:50 +0000 X-ASF-Spam-Status: No, hits=0.8 required=10.0 tests=AWL,FREEMAIL_FROM,HTML_MESSAGE,RCVD_IN_DNSWL_NONE,SPF_PASS,T_TO_NO_BRKTS_FREEMAIL X-Spam-Check-By: apache.org Received-SPF: pass (athena.apache.org: domain of masoodmortazavi@gmail.com designates 209.85.222.191 as permitted sender) Received: from [209.85.222.191] (HELO mail-pz0-f191.google.com) (209.85.222.191) by apache.org (qpsmtpd/0.29) with ESMTP; Tue, 20 Apr 2010 17:35:46 +0000 Received: by pzk29 with SMTP id 29so5366056pzk.29 for ; Tue, 20 Apr 2010 10:35:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:mime-version:received:in-reply-to:references :date:received:message-id:subject:from:to:cc:content-type; bh=qZ+8h74icx48sJv6PCh7iC9tR+tX6mdSXcz9JAGJ8yQ=; b=qBBnKZIJHe4qAid5JXpcgPYG0iFVbMZFEsnNHaG38qvKWMJ1/att3FIzimfPFPFrEn Z8WNJgKsloproPBdMMupXrHzYIJHLDfVGQFhEBDJb21bzjpbRAhkHyeiD3vtix/tRvrp nppr4U7OGqjDSs2XtOc/Q9mlWUeipBhGUSimk= DomainKey-Signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :cc:content-type; b=U5S7ZcIGBBvKYZzMp+Ei8/JklmXWKIJ4ZUph0mh00GU6cXgghxXtSCmPa918k31EEf BxA2Kend4/hQDFiiZQpiKanBQ33O57aQEDM1iFe9Ov2yzsFx7wMN2LHs1Ly+GKWfdZuy VMwHkVDeeCfWmrJBV6T//6wOB5v+UMfnrhmDI= MIME-Version: 1.0 Received: by 10.141.50.20 with HTTP; Tue, 20 Apr 2010 10:35:26 -0700 (PDT) In-Reply-To: References: <33FDEB0CE2F65F41A4CF8769247BB3668DC5DAA602@EXVMBX016-3.exch016.msoutlookonline.net> Date: Tue, 20 Apr 2010 10:35:26 -0700 Received: by 10.141.91.3 with SMTP id t3mr1855887rvl.191.1271784926194; Tue, 20 Apr 2010 10:35:26 -0700 (PDT) Message-ID: Subject: Re: 0.6 insert performance .... Re: [RELEASE] 0.6.1 From: Masood Mortazavi To: user@cassandra.apache.org Cc: "dev@cassandra.apache.org" Content-Type: multipart/alternative; boundary=000e0cd11092c1ad890484ae8080 --000e0cd11092c1ad890484ae8080 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable You're welcome Schubert. I look forward to any new results you may come up with. { It would also be interesting, when you run your tests again, to look at the GC logs and see to what extent https://issues.apache.org/jira/browse/CASSANDRA-896 is the culprit for what you will see. Identifying any other bottlenecks would be good, too. By the way, it is always good to list what JVM you're using. } On Mon, Apr 19, 2010 at 8:18 PM, Schubert Zhang wrote: > Since the scale of GC graph in the slides is different from the throughpu= t > ones. I will do another test for this issue. > Thanks for your advices, Masood and Jonathan. > > --------------- > Here, i just post my cossandra.in.sh. > JVM_OPTS=3D" \ > -ea \ > -Xms128M \ > -Xmx6G \ > -XX:TargetSurvivorRatio=3D90 \ > -XX:+AggressiveOpts \ > -XX:+UseParNewGC \ > -XX:+UseConcMarkSweepGC \ > -XX:+CMSParallelRemarkEnabled \ > -XX:SurvivorRatio=3D128 \ > -XX:MaxTenuringThreshold=3D0 \ > -Dcom.sun.management.jmxremote.port=3D8081 \ > -Dcom.sun.management.jmxremote.ssl=3Dfalse \ > -Dcom.sun.management.jmxremote.authenticate=3Dfalse" > > > On Tue, Apr 20, 2010 at 5:46 AM, Masood Mortazavi < > masoodmortazavi@gmail.com> wrote: > >> Minimizing GC pauses or minimizing time slots allocated to GC pauses -- >> either through configuration or re-implementations of garbage collection >> "bottlenecks" (i.e. object-generation "bottlenecks") -- seem to be the >> immediate approach. (Other approaches appear to be more intrusive.) >> At code level, using the GC logs, one can investigate further. There may >> be places were some object recycling can make some larger difference. >> Trying this first will probably bear more immediate fruit. >> >> - m. >> >> >> On Mon, Apr 19, 2010 at 9:11 AM, Daniel Kluesing wrote: >> >>> We see this behavior as well with 0.6, heap usage graphs look almost >>> identical. The GC is a noticeable bottleneck, we=92ve tried jdku19 and = jrockit >>> vm=92s. It basically kills any kind of soft real time behavior. >>> >>> >>> >>> *From:* Masood Mortazavi [mailto:masoodmortazavi@gmail.com] >>> *Sent:* Monday, April 19, 2010 4:15 AM >>> *To:* user@cassandra.apache.org; dev@cassandra.apache.org >>> *Subject:* 0.6 insert performance .... Re: [RELEASE] 0.6.1 >>> >>> >>> >>> I wonder if anyone can use: >>> >>> * Add logging of GC activity (CASSANDRA-813) >>> to confirm this: >>> >>> http://www.slideshare.net/schubertzhang/cassandra-060-insert-throughput >>> >>> - m. >>> >>> On Sun, Apr 18, 2010 at 6:58 PM, Eric Evans >>> wrote: >>> >>> >>> Hot on the trails of 0.6.0 comes our latest, 0.6.1. This stable point >>> release contains a number of important bugfixes[1] and is a painless >>> upgrade from 0.6.0. >>> >>> Enjoy! >>> >>> [1]: http://bit.ly/9NqwAb (changelog) >>> >>> -- >>> Eric Evans >>> eevans@rackspace.com >>> >>> >>> >> >> > --000e0cd11092c1ad890484ae8080 Content-Type: text/html; charset=windows-1252 Content-Transfer-Encoding: quoted-printable You're welcome Schubert.
I look forward to any new results you may c= ome up with.

{ It would also be interesting, when you run your tests= again, to look at the GC logs and see to what extent https://issues.apache.org/jira/b= rowse/CASSANDRA-896 is the culprit for what you will see. Identifying a= ny other bottlenecks would be good, too. By the way, it is always good to l= ist what JVM you're using. }

On Mon, Apr 19, 2010 at 8:18 PM, Schubert Zh= ang <zsongbo@gmai= l.com> wrote:
Since the scale of GC graph in the slides is different from the throughput = ones. I will do another test for this issue.
Thanks for your advices, Ma= sood and Jonathan.

---------------
Here, i just post my cossandra.in.sh.
JVM_OPTS=3D" \
=A0=A0=A0=A0=A0=A0=A0 -ea \
=A0=A0=A0=A0=A0=A0=A0= -Xms128M \
=A0=A0=A0=A0=A0=A0=A0 -Xmx6G \
=A0=A0=A0=A0=A0=A0=A0 -XX:= TargetSurvivorRatio=3D90 \
=A0=A0=A0=A0=A0=A0=A0 -XX:+AggressiveOpts \=A0=A0=A0=A0=A0=A0=A0 -XX:+UseParNewGC \
=A0=A0=A0=A0=A0=A0=A0 -XX:+Us= eConcMarkSweepGC \
=A0=A0=A0=A0=A0=A0=A0 -XX:+CMSParallelRemarkEnabled \
=A0=A0=A0=A0=A0=A0=A0 -XX:SurvivorRatio=3D128 \
=A0=A0=A0=A0=A0=A0=A0 -X= X:MaxTenuringThreshold=3D0 \
=A0=A0=A0=A0=A0=A0=A0 -Dcom.sun.management.= jmxremote.port=3D8081 \
=A0=A0=A0=A0=A0=A0=A0 -Dcom.sun.management.jmxre= mote.ssl=3Dfalse \
=A0=A0=A0=A0=A0=A0=A0 -Dcom.sun.management.jmxremote.= authenticate=3Dfalse"


On Tue, Apr 20, 2010 at 5:46 AM, Masood Mort= azavi <masoodmortazavi@gmail.com> wrote:
Minimizing GC pauses or minimizing time slots allocated to GC pauses -- eit= her through configuration or re-implementations of garbage collection "= ;bottlenecks" (i.e. object-generation "bottlenecks") -- seem= to be the immediate approach. (Other approaches appear to be more intrusiv= e.)
At code level, using the GC logs,=20 one can investigate further. There may be places were some object=20 recycling can make some larger difference.=A0
Trying this first will pr= obably bear more immediate fruit.

- m.


On Mon, Apr 19, 2= 010 at 9:11 AM, Daniel Kluesing <dk@bluekai.com> wrote:

We see this behavior as well with 0.6, heap usage graphs look almost identical. The GC is a noticeable bottleneck, we=92ve tried jdku19 and jrockit vm=92s. It basically kills any kind of soft real time behavior.=

=A0

From:= Masood Mortazavi [mailto:maso= odmortazavi@gmail.com]
Sent: Monday, April 19, 2010 4:15 AM
To: u= ser@cassandra.apache.org; dev@cassandra.apache.org
Subject: 0.6 insert performance .... Re: [RELEASE] 0.6.1

=A0

I wonder if any= one can use:


=A0* Add logging of GC activity (CASSANDRA-813)
to confirm this:
=A0 http://www.slideshare.net/schubertzhang/cass= andra-060-insert-throughput

- m.




--000e0cd11092c1ad890484ae8080--