Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (athena.apache.org: domain of dvankley@salesforce.com
 designates 209.85.218.47 as permitted sender)
MIME-Version: 1.0
Date: Fri, 24 Oct 2014 13:35:05 -0400
Message-ID: 
 <CAETT45gEpp_5MWeQ-xnih3dRHDrg1LQR3hAK4ygPwU_o0pCMxg@mail.gmail.com>
Subject: Re: Intermittent long application pauses on nodes
From: Dan van Kley <dvankley@salesforce.com>
To: user@cassandra.apache.org
Content-Type: multipart/alternative; boundary=089e0111de2afd135005062e9963

--089e0111de2afd135005062e9963
Content-Type: text/plain; charset=UTF-8

I'm also curious to know if this was ever resolved or if there's any other
recommended steps to take to continue to track it down. I'm seeing the same
issue in our production cluster, which is running Cassandra 2.0.10 and JVM
1.7u71, using the CMS collector. Just as described above, the issue is long
"Total time for which application threads were stopped" pauses that are not
a direct result of GC pauses (ParNew, initial mark or remark). When I
enabled the safepoint logging I saw the same result, long "sync" pause
times with short spin and block times, usually with the "RevokeBias"
description. We're seeing pause times sometimes in excess of 10 seconds, so
it's a pretty debilitating issue. Our machines are not swapping (or even
close to it) or having other load issues when these pauses occur. Any ideas
would be very appreciated. Thanks!

--089e0111de2afd135005062e9963
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">I&#39;m also curious to know if this was ever resolved or =
if there&#39;s any other recommended steps to take to continue to track it =
down. I&#39;m seeing the same issue in our production cluster, which is run=
ning Cassandra 2.0.10 and JVM 1.7u71, using the CMS collector. Just as desc=
ribed above, the issue is long &quot;Total time for which application threa=
ds were stopped&quot; pauses that are not a direct result of GC pauses (Par=
New, initial mark or remark). When I enabled the safepoint logging I saw th=
e same result, long &quot;sync&quot; pause times with short spin and block =
times, usually with the &quot;RevokeBias&quot; description. We&#39;re seein=
g pause times sometimes in excess of 10 seconds, so it&#39;s a pretty debil=
itating issue. Our machines are not swapping (or even close to it) or havin=
g other load issues when these pauses occur. Any ideas would be very apprec=
iated. Thanks!</div>

--089e0111de2afd135005062e9963--