[OpenIndiana-discuss] Openindiana GlusterFS support

[Jim Klimov]

On 2013-06-18 22:03, Heiko L. wrote:
> Hallo Alberto, Randy
>
>> Can you please comment which production problems have you found using your implementation?.
>
> My tests started on 30.04.
> Cluster is not in production yet.
> The next step is the installation of two VMs for test of performance.
> The main problem is rule no. 5 yet. [1]
>
> Regards Heiko
> (sorry for my english)
>
> [1] http://lists.nongnu.org/archive/html/gluster-devel/2013-06/msg00037.html

Hello, sorry to bring up a dated thread, but I wondered if
there are any news or caveats with OI+GlusterFS at the moment?

I was looking at some options for replicating data among my
family's HomeNAS rigs based on OpenIndiana, and it may be that
a subsystem like Gluster could come into equation (and maybe
something like clustered Alfresco on top for user-access over
several protocols [1]).

Regarding your question on co-existence of two nfs daemons
(Solaris and glusterd, [2] and I think your link was wrong
above) my instant reaction today was "Doesn't gluster run
in a local zone?" Did you try that, or work around in some
other way?

And... what was the purpose of those two VM's? Just some
applied load to see how the storage subsystem behaves? :)

Finally, am I correct that such setups basically use ZFS for
its volume management - redundancy and checksumming features,
dedup and caching, while other ZFS things like snapshots or
native NFS/CIFS/iSCSI integration remain overboard and are
re-published by respective application software (gluster, or
in my case even alfresco, etc.)? Even the local system file
access won't be to ZFS but rather via the glusterfs client
and its own POSIX layer first?

Blogs from different years are misleading... there is now a
glusterfs client that works with illumos/OI/Solaris, right?
The stack does not anymore require any Linux nodes?

Thanks,
//Jim Klimov

[1] 
http://blog.abstractive.ca/2012/07/implementing-a-complete-high-availability-alfresco-solution-using-open-source-technologies/

[2] 
http://www2.fh-lausitz.de/launic/comp/sol/130430.osol.diag_make_glusterfs/130525.xvm.gluster.nfs.overview.txt