Steven Whitehouse wrote:
Well, can't we (the Redhat/Centos fanboys) expect a critical Clustered
filesystem like GFS2 (Which supports over 16TB on a 64-bit bit systems
at least) take a leaf or two from () ZFS on this issue?
I'm not quite sure which feature you are suggesting that we take, but
I'd be surprised that if the start of a ZFS filesystem were to be
overwritten that it could be easily reconstructed.
ZFS is paranoid about metadata integrity to a degree which some might
regard as obsessional. There are at least 2 copies of everything and 4
for anything critical.
Of course it does have the advantage of not being a cluster filesystem
and having built in raid modes which makes raid6 look a bit careless,
along with online fsck and the ability to detect silent disk errors
(this is important given that it's statistically likely that a 2Tb
512bytes/sector drive will have undetectable errors not picked up by
onboard ECC a couple of times a year)
Which is not to say that we couldn't usefully learn a few lessons from
what other filesystems are doing, but only that I'm not sure that it
would help for this particular issue.
Bearing in mind that ZFS code is cddl, not GPLv2 (or v3), I believe that
it'd be worth looking at the design principles.
I've abused my test boxes in ways which have irreversably corrupted
every other filesystem and the worst ZFS has ever done is take the FS
offline.
I really wish there was some way of bringing ZFS into the RHEL fold (ie,
as a supported FS) as in my opinion it beats the pants off XFS, Ext3/4
or btrfs.
Alan
--
Linux-cluster mailing list
Linux-cluster@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/linux-cluster
