<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<br>
<br>
<div class="moz-cite-prefix">On 05/29/2015 12:34 PM, Pedro Oriani
wrote:<br>
</div>
<blockquote cite="mid:SNT146-W1166165E8D46A6CAB5E54AA0C90@phx.gbl"
type="cite">
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style>
<div dir="ltr">Hi <span style="color: rgb(68, 68, 68); font-size:
15px; line-height: 21.2999992370605px; background-color:
rgb(255, 255, 255);">Pranith,</span>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;"><br>
</span></font></div>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;">it's for sure related to
a replication / healing task, because occurses when you
create a new replicated brick or when you bring back
online an old one.</span></font></div>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;">The problem is that the
cpu load on the online brick is so high that I cannot do
normal operations.</span></font></div>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;">In my case when a
replication / healing occurs, the cluster cannot serve
content.</span></font></div>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;">I'm asking if there is a
way to limit cpu usage in this case, or set a less
aggressive mode, because otherwise I have to rethink the
image repository.</span></font></div>
</div>
</blockquote>
<font color="#444444">Disable self-heal. I see that you already did
that for self-heal daemon. Lets do that even for mounts.</font><br>
gluster volume set <volname> cluster.entry-self-heal off<br>
<br>
Let me know how that goes.<br>
<br>
Pranith<br>
<blockquote cite="mid:SNT146-W1166165E8D46A6CAB5E54AA0C90@phx.gbl"
type="cite">
<div dir="ltr">
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;"><br>
</span></font></div>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;">thanks,</span></font></div>
<div><font color="#444444"><span style="font-size: 15px;
line-height: 21.2999992370605px;">Pedro</span></font></div>
<div><br>
<div>
<hr id="stopSpelling">Date: Fri, 29 May 2015 11:14:29 +0530<br>
From: <a class="moz-txt-link-abbreviated" href="mailto:pkarampu@redhat.com">pkarampu@redhat.com</a><br>
To: <a class="moz-txt-link-abbreviated" href="mailto:sgunfio@hotmail.com">sgunfio@hotmail.com</a>; <a class="moz-txt-link-abbreviated" href="mailto:gluster-users@gluster.org">gluster-users@gluster.org</a><br>
Subject: Re: [Gluster-users] 100% cpu on brick replication<br>
<br>
<br>
<br>
<div class="ecxmoz-cite-prefix">On 05/27/2015 08:48 PM,
Pedro Oriani wrote:<br>
</div>
<blockquote
cite="mid:SNT146-W71009E41D00B73F9D90120A0CB0@phx.gbl">
<style><!--
.ExternalClass .ecxhmmessage P {
padding:0px;
}
.ExternalClass body.ecxhmmessage {
font-size:12pt;
font-family:Calibri;
}
--></style>
<div dir="ltr">Hi All,
<div>I'm writing because I'm experiecing an issue with
gluster's replication feature.</div>
<div>I've a brick on srv1 with about 2TB of mixed side
files, ranging from 10k a 300k</div>
<div>When I add a new replication brick on srv2, the
glusterfs process take all the cpu.</div>
<div>This is unsuitable because the volume is not
responding at normal r/w queries.</div>
<div><br>
</div>
<div>Glusterfs version is 3.7.0</div>
</div>
</blockquote>
Is it because of self-heals? Was the brick offline until
then?<br>
<br>
Pranith<br>
<blockquote
cite="mid:SNT146-W71009E41D00B73F9D90120A0CB0@phx.gbl">
<div dir="ltr">
<div><br>
</div>
<div>the underlaying volume is xfs.</div>
<div><br>
</div>
<div><br>
</div>
<div>
<div>Volume Name: vol1</div>
<div>Type: Replicate</div>
<div>Volume ID: </div>
<div>Status: Started</div>
<div>Number of Bricks: 1 x 2 = 2</div>
<div>Transport-type: tcp</div>
<div>Bricks:</div>
<div>Brick1:
172.16.0.1:/data/glusterfs/vol1/brick1/brick</div>
<div>Brick2:
172.16.0.2:/data/glusterfs/vol1/brick1/brick</div>
<div>Options Reconfigured:</div>
<div>performance.cache-size: 1gb</div>
<div>cluster.self-heal-daemon: off</div>
<div>cluster.data-self-heal-algorithm: full</div>
<div>cluster.metadata-self-heal: off</div>
<div>performance.cache-max-file-size: 2MB</div>
<div>performance.cache-refresh-timeout: 1</div>
<div>performance.stat-prefetch: off</div>
<div>performance.read-ahead: on</div>
<div>performance.quick-read: off</div>
<div>performance.write-behind-window-size: 4MB</div>
<div>performance.flush-behind: on</div>
<div>performance.write-behind: on</div>
<div>performance.io-thread-count: 32</div>
<div>performance.io-cache: on</div>
<div>network.ping-timeout: 2</div>
<div>nfs.addr-namelookup: off</div>
<div>performance.strict-write-ordering: on</div>
</div>
<div><br>
</div>
<div><br>
</div>
<div>there is any parameter or hint that I can follow to
limit cpu occupation to grant a replication with few
lag on normal operations ?</div>
<div><br>
</div>
<div>thank </div>
</div>
<br>
<fieldset class="ecxmimeAttachmentHeader"></fieldset>
<br>
<pre>_______________________________________________
Gluster-users mailing list
<a moz-do-not-send="true" class="ecxmoz-txt-link-abbreviated" href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a>
<a moz-do-not-send="true" class="ecxmoz-txt-link-freetext" href="http://www.gluster.org/mailman/listinfo/gluster-users" target="_blank">http://www.gluster.org/mailman/listinfo/gluster-users</a></pre>
</blockquote>
<br>
</div>
</div>
</div>
</blockquote>
<br>
</body>
</html>