<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
I'll be interested to hear how this works for you. THanks for the
update.<br>
<br>
-Collin<br>
<br>
Dai, Manhong wrote:
<blockquote
cite="mid:82343BC0ECA21B499F8065BB542B008907791D@shops1075.mbni.med.umich.edu"
type="cite">
<meta http-equiv="Content-Type" content="text/html; ">
<meta name="Generator"
content="MS Exchange Server version 6.5.7653.38">
<title>Re: help, glusterfs test caused very high tcp segment
retransmission rate (Collin Douglas)</title>
<!-- Converted from text/plain format -->
<br>
<p><font size="2">Hi Collin,<br>
<br>
Thanks a lot for your help. The interconnect is Gigabit Ethernet.<br>
<br>
I am guessing that it is caused by the router. Computing nodes that
connect to this router always retransmit 5% of segments under heavy
load. My test on other router has lower ratio, around 0.3%. I am about
to replace the router and perform another test.<br>
<br>
Anyway this problem is not related to glusterfs at all. Although I do
have another question that is about glusterfs, which I will send in a
separated email.<br>
<br>
Best,<br>
Manhong<br>
<br>
Date: Mon, 25 Aug 2008 08:41:25 -0500<br>
From: Collin Douglas <a class="moz-txt-link-rfc2396E" href="mailto:cbd@adfitech.com"><cbd@adfitech.com></a><br>
Subject: Re: [Gluster-users] help, glusterfs test caused very high tcp<br>
segment retransmission rate<br>
Cc: <a class="moz-txt-link-abbreviated" href="mailto:gluster-users@gluster.org">gluster-users@gluster.org</a><br>
Message-ID: <a class="moz-txt-link-rfc2396E" href="mailto:48B2B685.2050603@adfitech.com"><48B2B685.2050603@adfitech.com></a><br>
Content-Type: text/plain; charset="ISO-8859-1"<br>
<br>
-----BEGIN PGP SIGNED MESSAGE-----<br>
Hash: SHA256<br>
<br>
This sounds a lot like a physical layer issue -- most notably a<br>
speed/duplex issue. Any time I've seen a large number of retransmits<br>
occur during heavy load, this is what I suspect.<br>
<br>
That said, I don't know what type of interconnect you're using. Tell us<br>
more about your configuration.<br>
<br>
- -Collin<br>
<br>
Dai, Manhong wrote:<br>
><br>
><br>
> Hi,<br>
><br>
> <br>
> I found the aggregated IO speed is only about 100MB/s on 4
Giga-bit<br>
> Brick. This test is done over 12 computing nodes with command dd<br>
> if=/dev/zero of=bar bs=1048576 count=20480. Because our brick has
very<br>
> fast local IO speed, the problem could be network.<br>
><br>
><br>
> Then I found computing nodes got too many retransmited segments<br>
> during test according to netstat -st. The retransmission ratio is<br>
> about 5%, but brick node has a normally low transmission rate.<br>
><br>
><br>
> Could some shed some light?<br>
><br>
> Best,<br>
> Manhong<br>
</font>
</p>
<pre wrap="">
<hr size="4" width="90%">
_______________________________________________
Gluster-users mailing list
<a class="moz-txt-link-abbreviated" href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a>
<a class="moz-txt-link-freetext" href="http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users">http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users</a>
</pre>
</blockquote>
</body>
</html>