[aklug] Re: Recommendations for DRBD + (GFS | Lustre)

From: Shane R. Spencer <shane@bogomip.com>
Date: Sun Jan 18 2009 - 11:48:59 AKST

There's lots of ways to do what you want. You can even use NBD (Network
Block Device) and have one machine responsible for making sure the array
is clean and up to date. DRBD supports Master-Slave and Master-Master
(I think it does anyways, and therefore Master-Master-Slave if you add a
new drbd on top of the existing one). When using Master-Master you have
to use a global lock file based filesystem. But Master-Master can be a
pain to administrate.
Using GFS2 you can format and use any "redundant" block device or
network block device on your network and have a good global lock system
to make sure write allocation works correctly.

GlusterFS is fun, easy to install, easy to maintain. Make sure you use
locks if you're writing tons of data. Just export some directories, set
up how you want clients to connect and how they perceive the cluster and
go with it. It uses the distributed hash table approach toward storing
data, however unlike most DHT's it is incredibly flexible and supports
several redundancy methods. Each "volume" is stackable to be able to
change IO parameters as well as make mirrors of mirrors of distributed
maps of penguins. You know what I'm talkin' about. It's like NFS on roids.

I've set up firewire drives that sit between two servers and use GFS2
perfectly. Not very hot-redundant but it worked better than DRBD :)
Same concept goes for iSCSI, AoE, NBD.

Shane

Jeremy Austin wrote:
> Looking for the wisdom of the crowd:
> I've been running DRBD on (small) clusters of servers for a few years. (DRBD
> is a distributed RAID block device, similar to md -- software raid, but with
> hard drives spread out over multiple servers rather than just one.)
> Until now I've been using the Primary/Secondary mode, where only one half of
> the DRBD can write to disk at any one time.
>
> I'd like to switch to a clustered filesystem, and I see that DRBD has
> support for GFS. Does anyone in the group have experience with GFS?
>
> Another filesystem alternative appears to be Lustre, which seems (at first
> glance) to provide very high performance, but be much more complicated.
>
> In general I don't need very high performance; I'm running a few web sites,
> file servers, and VMware machines, supporting users connected via DSL. I'm
> more interested in the high availability of DRBD (and that works very well),
> potentially coupled with the higher overall performance of a clustered
> filesystem.
>
> My goal is a 3-server cluster, with 3 arrays (one per 'pair' of servers.)
> Should I try DRBD+GFS, DRBD+Lustre, or stick with plain DRBD & ext3, as I've
> been doing?
>
> Thanks as always,
> Jeremy Austin
> IT Administrator
> Whitestone, Alaska
>
>
> ---------
> To unsubscribe, send email to <aklug-request@aklug.org>
> with 'unsubscribe' in the message body.
>

-- Attached file included as plaintext by Ecartis --
-- File: signature.asc
-- Desc: OpenPGP digital signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAklzlb4ACgkQXK/vGhypreI7GwCfeOBOl/+V2j4mQIVDpLgW6x3h
V9QAnjOe843KyZQpZn6/PVxKzhfNEBO1
=vZsE
-----END PGP SIGNATURE-----

---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Sun Jan 18 11:49:13 2009

This archive was generated by hypermail 2.1.8 : Sun Jan 18 2009 - 11:49:13 AKST