Hi, all,
I've searched through the archives and saw that this has come up before,
but it was not clear what the best way was to accomplish this. I'm
currently on 3.2.3 and would like to convert a two-brick distribute
volume to a four-brick disribute-replicate volume. Would editing the
vol file directly by encapsulating the bricks I want to replicate in the
replicate translator, then encapsulating that in the distribute
translator be the way to go, then ls -RA to trigger a self-heal ? Is
there a way to do this from the gluster console command?
Thanks,
-Brian
Brian Smith
Senior Systems Administrator
IT Research Computing, University of South Florida
4202 E. Fowler Ave. ENB308
Office Phone: +1 813 974-1467
Organization URL: http://rc.usf.edu
On 10/27/2011 02:00 AM, Mohammed Junaid wrote:
Hi Daniel,
However, I've been trying to setup an extra swift proxy to offload some of
the CPU usage I've been experiencing in our tests and I can't really find a
way to do so.
What's the correct way to proceed, with gluster's object-storage? Anyone
else has done it?
To increase the proxy server count, edit the file
"/etc/gluster-object/proxy-server.conf and add the line
workers =<count>
Following Swift's documentation I get stuck at copying the ring files (do
they exist in gluster's object-storage?). Then I've thought I could simply
setup gluster object-storage in a proxy machine and somehow redirect the
requests from the proxy to the storage machine, but I'm failing miserably
that way as well.
For using gluster object-storage, there is no need to generate the ring
files. Please follow the gluster object-storage documentation to setup the
machines.
Thanks,
Junaid
_______________________________________________
Gluster-users mailing list
Gluster-users@gluster.org
http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
_______________________________________________
Gluster-users mailing list
Gluster-users@gluster.org
http://gluster.org/cgi-bin/mailman/listinfo/gluster-users
