2

I'm thinking of migrating my server park to ZFS(not really), but this live demo makes me suspicious.

My question is:

  1. Is what we see here normal behavior for ZFS?

http://www.youtube.com/watch?v=CN6iDzesEs0&feature=player_embedded#t=4m40s

Any way, this is a Kodak moment. The speaker seems not to expect what he see during the first 2 seconds. Hilarious!

1
  • There is a reason why the advise is "Never do live demo's" :-)
    – Vagnerr
    May 29, 2009 at 9:13

2 Answers 2

5

ZFS had it's fair share of issues, but it gets very much attention at Sun. I wouldn't be surprised if the problems at hand are already solved. And drive failure like shown in that case is something every RAID-Controller/Filesystem loathes. I've seen more than one RAID-Controller causing the OS to panic/hang due to drive electronics talking dirty to it (and that was without sledgehammers ;) ).

On the other side I don't think that the question is if "ZFS is any good" but "When will it take over the storage sector".

All the positive things about ZFS (VERY easy block device management, lean provisioning of filesystems, variable block sizes, non-dependency on battery backed write caches for safe operation, snapshots "for free", etc.) are going to be noticed and requested by the customers.

And this is probably why every storage vendor that is not NetApp is slowly becoming nervous these days (they're stuck with their archaic storage designs since a very long time) and Oracle sponsors btrfs development on Linux (though this has become some sort of a moot point since they've vowed to acquire Sun...).

0

Yeh, I noticed that too.

I suspect that it is down to the disks being USB attached and the USB controllers not being smart enough to report that the disk has failed. As soon as the usb cable is disconnected the disk subsystem springs back into life.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .