Hi Matthew,<br><br>please find the inlined comments.<br><br><div class="gmail_quote">On Tue, Apr 7, 2009 at 5:38 AM, Matthew Wilkins <span dir="ltr"><<a href="mailto:daibutsu@gmail.com">daibutsu@gmail.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi,<br>
<br>
I am currently running glusterfs-1.3.10 in a NUFA situation using the<br>
unify translator, nufa scheduler, and an AFR'ed namespace. A<br>
config from one of my servers is below if you want the full<br>
details. I want to upgrade to 2.0.0rc7 and use the nufa<br>
translator instead of unify. The nufa cluster translator sounds like<br>
the better option. I have some questions for you helpful people!<br>
<br>
1. Is it possible to do such an upgrade? I was thinking I would<br>
umount the fs and stop the gluster daemons. Upgrade to 2.0.0rc7 and<br>
put the new config files in. Mount up the fs again. Note the<br>
namespace will have disappeared because it isn't necessary in version<br>
2 right? So what will happen now? Will there be a lot of<br>
self-healing necessary? Can that just happen as people use the<br>
gluster? (btw, total size is 32T, used size is about 15T).<br>
<br>
2. I have a sneaking suspicion that the answer to 1 is 'No it wont<br>
work'. Either way I am wondering how the distributed translator<br>
works. The nufa translator a special case of the distributed<br>
translator correct? (but is must have some bias towards the local<br>
server?). Anyway, how do they work? When a client wants to find a<br>
file a hash is done on the filename, somehow that maps to a particular<br>
server, then the client asks that server for the file? Are any<br>
extended attributes set on the file?</blockquote><div><br>yes, files are mapped to servers based on their hash value. DHT does not use any extended attributes.<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
<br>
3. One of my servers has two bricks. The nufa example at<br>
<a href="http://gluster.org/docs/index.php/NUFA_with_single_process" target="_blank">http://gluster.org/docs/index.php/NUFA_with_single_process</a><br>
doesn't show me what to do. It has two examples; the first when each node<br>
has one brick, and another where nodes have more than one brick but<br>
nufa is not used, rather unify. So how can I used the nufa translator<br>
when one or more nodes contribute more than one brick? I was thinking<br>
something like a server side unify, then nufa on top, but I'm not sure<br>
of the syntax. If it isn't possible it isn't the end of the world<br>
(the second brick isn't that big).</blockquote><div><br>export each brick separately and have the client protocols corresponding to each of the exported brick as children of nufa.<br> </div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
<br>
4. At the end of my new config for version 2 I have the following:<br>
<br>
volume nufa<br>
type cluster/nufa<br>
option local-volume-name `hostname`<br>
subvolumes tur-awc1 tur-awc2 tur-awc3<br>
<br>
volume writebehind<br>
type performance/write-behind<br>
option page-size 128KB<br>
option cache-size 1MB<br>
subvolumes nufa<br>
end-volume<br>
<br>
volume ra<br>
end-volume<br>
<br>
Is that the correct order for these performance translators, it isn't<br>
obvious to me if the read-ahead should be before or after write-behind<br>
(all the examples I have seen have io-cache at the end though). Does<br>
the order matter?</blockquote><div></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><br>
<br>
Thank you very much for any help. If you can only help with one<br>
question I would still very much appreciate it.<br>
<br>
Matt<br>
<br>
Here is my current config:<br>
<br>
volume brick0<br>
type storage/posix<br>
option directory /export/brick0<br>
end-volume<br>
<br>
volume iot-0<br>
type performance/io-threads<br>
subvolumes brick0<br>
end-volume<br>
<br>
volume brick1<br>
type storage/posix<br>
option directory /export/brick1<br>
end-volume<br>
<br>
volume iot-1<br>
type performance/io-threads<br>
subvolumes brick1<br>
end-volume<br>
<br>
volume brick-ns<br>
type storage/posix<br>
option directory /export/brick-ns<br>
end-volume<br>
<br>
volume iot-ns<br>
type performance/io-threads<br>
subvolumes brick-ns<br>
end-volume<br>
<br>
volume server<br>
type protocol/server<br>
subvolumes iot-0 iot-1 iot-ns<br>
option transport-type tcp/server # For TCP/IP transport<br>
option auth.ip.iot-0.allow *<br>
option auth.ip.iot-1.allow *<br>
option auth.ip.iot-ns.allow *<br>
end-volume<br>
<br>
<br>
volume client-tardis-0<br>
type protocol/client<br>
option transport-type tcp/client<br>
option remote-host tardis<br>
option remote-subvolume iot-0<br>
end-volume<br>
<br>
volume client-orac-0<br>
type protocol/client<br>
option transport-type tcp/client<br>
option remote-host orac<br>
option remote-subvolume iot-0<br>
end-volume<br>
<br>
volume client-orac-ns<br>
type protocol/client<br>
option transport-type tcp/client<br>
option remote-host orac<br>
option remote-subvolume iot-ns<br>
end-volume<br>
<br>
volume afr-ns<br>
type cluster/afr<br>
subvolumes iot-ns client-orac-ns<br>
end-volume<br>
<br>
volume unify<br>
type cluster/unify<br>
option namespace afr-ns<br>
option scheduler nufa<br>
option nufa.local-volume-name iot-0,iot-1<br>
option nufa.limits.min-free-disk 5%<br>
subvolumes iot-0 iot-1 client-orac-0 client-tardis-0<br>
end-volume<br>
<br>
volume ra<br>
type performance/read-ahead<br>
subvolumes unify<br>
end-volume<br>
<br>
volume ioc<br>
type performance/io-cache<br>
subvolumes ra<br>
end-volume<br>
<br>
_______________________________________________<br>
Gluster-users mailing list<br>
<a href="mailto:Gluster-users@gluster.org">Gluster-users@gluster.org</a><br>
<a href="http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users" target="_blank">http://zresearch.com/cgi-bin/mailman/listinfo/gluster-users</a><br>
</blockquote></div><br><br clear="all"><br>-- <br>Raghavendra G<br><br>