I think anything based on fuse is doomed. Without being in the
kernel, how do you make sure data is actually on disk on the server in
any sensible order? Meta-data can be made atomic but without being in
the kernel, how do you actually do that for bulk data? Your not even
guaranteed that data goes to the server/disk in the order you send it.
I''m sure glustre works fine in the usual, non-stress, simple case -
but then most things do. Its the cases where things start going wrong
(memory pressure on the nodes or server, congestion on the network,
failed hardware) that the real problems arise. With hundreds of nodes
and 30+ file servers, failures happen daily. That scares me with a
fuse based file system.
--
Dr Stuart Midgley
sdm900 at gmail.com
On 21/07/2009, at 1:00 AM, Jordan Mendler wrote:
> Not to continue an off-topic thread, but how big was your Gluster
> deployment? I am curious because we found it unusable with ~50TB and
> 5-10 million files (even though our goal was several hundred).
>
> Jordan
>
> Hi all,
>
> Regarding to the comparsion of Lustre with GlusterFS, i have the
> fallowing together with the attached file.
>
> I used this in a presentation at
> http://www.beliefproject.org/events/4th-belief-international-symposium
>
> It can be a start of comparsion. GlusterFS has a lower performance,
> but we can make a RAID 10 though the network, in the case that the
> glusterfs equivalent of OST are DAS( ie, internal disk, not a SAN
> lun). Also, apparently, there is no deadlock when the glusterfs node
> is simultaneously configured to be client and server.
>
> In my experince, the integration between most recent kernels with
> glusterfs and patches of Xen hypervisor works well. The same with
> Lustre is harder to do.
>
> Although, there is issues when trying to boot up a virtual machine
> image stored in glusterfs mount point.
>
> Also glusterfs has lower performance when overwriting files when
> compared with the write process.
>
> Apparently glusterfs does not stripe files between nodes and
> apparently there is a single file limit size asspciated to the space
> available in the gluster equivalent to OST.
>
> I am waiting for version 2 of Lustre.
>
> Best Regards.
> --
> Ettore Enrico Delfino Ligorio
> ettoredelfinoligorio at gmail.com
> 55-11-9145-6151
> _______________________________________________
> Lustre-discuss mailing list
> Lustre-discuss at lists.lustre.org
> http://lists.lustre.org/mailman/listinfo/lustre-discuss