Is it safe to run RAID0 on a replicate cluster?

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Hi everyone!

I've asked this question before, but I don't think my subject line got the right people's attention, so apologies for re-posting.

We've got a mail server with about 10,000 e-mail accounts, all sharing their Maildir mailboxes on our 3-node Gluster array configured as a replicated volume. The performance of this Gluster share is quite poor, and it takes a long time to fetch peoples' mail, especially when they're using webmail.

Every node in the Gluster array has their RAID array configured as RAID5, so I'd like to improve the performance on each node by changing that to RAID0 instead. I think that that Gluster's data replication is good enough to keep our data from being lost in the event of a disk failure, but I want other people's opinions on the matter before we go doing something so drastic. At the very least, I need someone's opinion on this particular subject before we do anything.

One other person has recommended some Gluster settings to improve Gluster performance for small files, but considering the IO and CPU overhead of Gluster, I'd have to wonder about its usefulness.

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
http://lists.gluster.org/mailman/listinfo/gluster-users

[Index of Archives]     [Gluster Development]     [Linux Filesytems Development]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux