<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#3333FF">
<font size="+1"><font face="monospace">This is going to be an
absolute game changer!<br>
<br>
I've been pulling my hair out trying to configure 4 boxes for
each location:<br>
2 for Proxmox VMs (Clustering), and 2 for data redundancy.<br>
<br>
Great work guys, I'm donating some more $$ today!<br>
<br>
-J<br>
</font></font><br>
Alain Péan said the following on 10/24/2011 06:05 AM:
<blockquote cite="mid:4EA53866.9070203@lpp.polytechnique.fr"
type="cite">Le 24/10/2011 11:32, Patryk Benderz a écrit :
<br>
<blockquote type="cite">[cut]
<br>
<blockquote type="cite">And a big difference to almost all
competitors: no single point of failure, due to pmxcfs.
<br>
</blockquote>
Could you elaborate particularly what are you referring to, when
<br>
speaking of "single point of failure"?
<br>
<br>
</blockquote>
<br>
Hi Patryk,
<br>
<br>
If I can answer for Martin... I think he is referring to what is
described in the roadmap :
<br>
<a class="moz-txt-link-freetext" href="http://pve.proxmox.com/wiki/Roadmap#Roadmap_for_2.x">http://pve.proxmox.com/wiki/Roadmap#Roadmap_for_2.x</a>
<br>
<br>
So, with pmxcfs, there is no more a single master node, that can
be a SPOF.
<br>
<br>
And I can add, that if 'Sheepdog' was configured (distributed
storage among the nodes), there will be no need for central
storage as iSCSI, or FC, that are also SPOF, even if they have
redundant Controller, Power Supply, multipath... etc (furthermore,
they are very expensive). See :
<br>
<a class="moz-txt-link-freetext" href="http://www.osrg.net/sheepdog/">http://www.osrg.net/sheepdog/</a>
<br>
<br>
Alain
<br>
<br>
</blockquote>
</body>
</html>