nbd is similar in function to iscsi; it exports block devices across a
network. Iscsi is probably a more mature than nbd by now. The limitation
is only one host can mount the nbd at a time. nbd reminds of ATA over
ethernet.
drbd replicates a block device across the network. Intended for high
availability and replication/snapshot backups. drbd was merged into
kernel 2.6.32 (SL 6 maybe?)
On Mon, 12 Apr 2010, Troy Dawson wrote:
> Hello,
> We are looking at puting drbd into SL5. I do not think our packaging is
> ready to go into SL 5.5, but I'd like to have it tested before putting
> it into contrib, and then we'll see whether we want to put it in SL 5.6.
>
> "DRBD is a block device which is designed to build high availability
> clusters. This is done by mirroring a whole block device via (a
> dedicated) network. You could see it as a network raid-1."
I know nothing about DRBD but isn't there already another implementation
of something very similar (NBD, ENBD etc)?
http://nbd.sourceforge.net/
http://www.it.uc3m.es/~ptb/nbd/
I know that some of the cluster vendors were taking the NBD/ENBD code and
improving it in various ways. Is this what DRBD is or does it do more
than md over NBD can?
[ Not that I've used NBD/ENBD either but I've been told that another
department are using it a lot as part of a backup solution (not that this
means much to us)... ]
--
