<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">02.02.2014 13:53, Valerio Pachera
пишет:</div>
<blockquote
cite="mid:CAHS0cb_SiK30omOypjQGd=fF157tARdTaRkyionURqdn6M7VVQ@mail.gmail.com"
type="cite">
<div dir="ltr"><br>
<div class="gmail_extra">2014-02-01 Maxim Terletskiy <span
dir="ltr"><<a moz-do-not-send="true"
href="mailto:terletskiy@emu.ru" target="_blank">terletskiy@emu.ru</a>></span>:<br>
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
the reason before found that "nodiratime" is root of the
evil. </blockquote>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">Hope
this information would be usefull for somebody.<br>
</blockquote>
<div><br>
Sure it is.<br>
</div>
<div>Has any body else been able to reproduce this problem?<br>
</div>
<div>What version of sheep are you running?<br>
</div>
<div><br>
</div>
</div>
</div>
</div>
</blockquote>
For me "nodiratime" problem arises with QEMU 1.6.1, 1.6.2, 1.7.0.
Using CentOS 6.5 x86_64, sheep version 0.7.6 with zookeeper, object
cache on ext4 partition, dual-nic with bonded ethernet for cluster
communication and IPoIB for IO. Sheep options I'm using on VM nodes:<br>
sheep -g -c
zookeeper:172.16.0.15:2181,172.31.200.249:2181,172.31.200.250:2181,timeout=30
-y 172.16.0.10 -i host=192.168.10.10,port=7001 -w
dir=/mnt/sheep_objcache,size=213000<br>
<br>
<blockquote
cite="mid:CAHS0cb_SiK30omOypjQGd=fF157tARdTaRkyionURqdn6M7VVQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
Now I'm curious about object cache failover. What happen
if volume with cache will fail? Will sheep and VMs live or
will they die?<br>
</blockquote>
<div><br>
</div>
<div>I'm also curious.<br>
</div>
<div>I think it's not handled as of now.<br>
</div>
<div>It would be good that, in case of I/O error from the
cache device, sheep disables cache (like a node md
unplug).<br>
</div>
</div>
</div>
</div>
</blockquote>
Morita Kazukata writes that VMs must survive. I will try to do some
tests.<br>
<blockquote
cite="mid:CAHS0cb_SiK30omOypjQGd=fF157tARdTaRkyionURqdn6M7VVQ@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">
<div> </div>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
(I found that with restart of sheep process VMs connected
to it dying)<br>
</blockquote>
<div><br>
</div>
<div>They shouldn't, but this depends on your qemu version.<br>
</div>
<div>Qemu supports auto-reconnect. As soon as sheep daemon
is available again, the vm will issue their I/O requests.<br>
</div>
<div> </div>
</div>
</div>
</div>
</blockquote>
Tried twice with QEMU 1.6.1. Both times result was negative.<br>
</body>
</html>