<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 07/23/2013 01:28 PM, Ziemowit
Pierzycki wrote:<br>
</div>
<blockquote
cite="mid:CAOiTGH=S+RHxb9O1F5j2ziYOaWf=wWysJuzYB1Qa0qMKwyMryw@mail.gmail.com"
type="cite">
<div dir="ltr">What is better? Is making a client part of the
cluster and then making the mount via "localhost":
<div style=""><br>
</div>
<div style=""># service glusterd start</div>
<div style="">
# gluster peer probe somehost</div>
<div style=""># mount -t glusterfs localhost:/volume /mnt/temp</div>
<div style=""><br>
</div>
<div style="">or:</div>
<div style=""><br>
</div>
<div style=""># mount -t glusterfs somehost:/volume /mnt/temp</div>
<div style=""><br>
</div>
<div style="">I guess with mounting it via somehost, one would
have to very picky to choose which host to pick as if it goes
down that mount may stop working. Am I correct? So what is
better?</div>
</div>
</blockquote>
<br>
Another option is to use round-robin dns to make a single hostname
point to all your gluster servers:
<meta http-equiv="content-type" content="text/html;
charset=ISO-8859-1">
<a
href="http://edwyseguru.wordpress.com/2012/01/09/using-rrdns-to-allow-mount-failover-with-glusterfs/">http://edwyseguru.wordpress.com/2012/01/09/using-rrdns-to-allow-mount-failover-with-glusterfs/</a>.
This gives you a single hostname to mount through, while still
working fine when some of your servers are down for maintenance.<br>
<br>
Pierre<br>
<br>
</body>
</html>