<div dir="ltr"><div><div><div>Hi Brian,<br><br></div>Thanks for your further explanation. Myself actually just proved that option 1 works for a totally damaged brick.<br></div> For option 2, is this UUID for the peer host? Because I found in my peer servers, each hosts's /var/lib/glusterd/<a href="http://glusterd.info" target="_blank">glusterd.info</a> holds the UUID stored from the directory /var/lib/glusterd/peers/ of other peer host.<br>
<br></div>Liang<br><div> <a href="http://glusterd.info" target="_blank"></a></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Jan 10, 2013 at 5:06 PM, Brian Candler <span dir="ltr"><<a href="mailto:B.Candler@pobox.com" target="_blank">B.Candler@pobox.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">On Thu, Jan 10, 2013 at 12:50:48PM -0500, Liang Ma wrote:<br>
> I assume to replace a failed replicate disk or node should be a<br>
> standard procedure, isn't it? I could find anything related to this in<br>
> the 3.3 manual.<br>
<br>
</div>You'd have thought so, wouldn't you :-(<br>
<br>
I know of two options.<br>
<br>
(1) If the server itself is OK, but the brick filesystem you're exporting<br>
from that server has died, then just stop glusterd, erase (mkfs) the<br>
filesystem which the brick is on, remount it, restart glusterd.<br>
<br>
After a few minutes, self-heal will kick in and copy the data for replicated<br>
volumes. At least, it did in a test setup I tried once.<br>
<br>
This of course assumes your data filesystem is separate from your OS<br>
filesystem, which I'd suggest is a good idea anyway.<br>
<br>
(2) If the whole server has died, or you have to re-install the OS from<br>
scratch, but the replacement server has the same hostname as the old one,<br>
then there's a different procedure.<br>
<br>
It was documented at<br>
<a href="http://gluster.org/community/documentation/index.php/Gluster_3.2:_Brick_Restoration_-_Replace_Crashed_Server" target="_blank">http://gluster.org/community/documentation/index.php/Gluster_3.2:_Brick_Restoration_-_Replace_Crashed_Server</a><br>
for glusterfs 3.2. It is almost the same for glusterfs 3.3, but the config<br>
directory has moved. It works if you change these two steps:<br>
<br>
grep server3 /var/lib/glusterd/peers/*<br>
<br>
echo UUID=... >/var/lib/glusterd/<a href="http://glusterd.info" target="_blank">glusterd.info</a><br>
<br>
HTH,<br>
<br>
Brian.<br>
</blockquote></div><br></div>