Re: [Gluster-devel] Problems with unify

[Amar S. Tumballi]

Hi Antonio,
 Unify's self-heal takes care of servers going down and coming up, and
checks for consistency. There was a bug introduced in '1.3.8pre4' which
caused unify self-heal to fail in this scenario. But its fixed in the
current '1.3.8pre5' release. Can you check with the latest pre release, and
see if the problem still remains?

Regards,
Amar


On Fri, Apr 4, 2008 at 3:45 AM, Antonio González <
address@hidden> wrote:

> Hello all,
>
>
>
> I have made some test with the unify translator to check the viability…The
> scenario is 4 machines, 2 storage servers, 1 server witch creates the
> unify
> translator, and 1 client. If the server 1 goes down and I "mkdir" in the
> client, when the server returns if I list the directory in the client, the
> directory is created in the server, but if I try to "cp" or "mv" a file/es
> directly over the directory (without using the "tab" to list the
> directory)
> an inconsistency occurs. The directory will be not created at the server
> and
> the client will not be able to read these files.
>
>
>
> In future releases, the self-heal of unify will be automatic??? I say,
> When
> a server goes down and returns after, a process checks automatically the
> consistency of namespace and child nodes.
>
>
>
> Thanks
>
>
>
> Antonio.
>
> _______________________________________________
> Gluster-devel mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/gluster-devel
>



-- 
Amar Tumballi
Gluster/GlusterFS Hacker
[bulde on #gluster/irc.gnu.org]
http://www.zresearch.com - Commoditizing Supercomputing and Superstorage!