[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [SAGE] nas boxes with snapshots
- To: sage-members@usenix.org
- Subject: Re: [SAGE] nas boxes with snapshots
- From: Phil Pennock <phil.pennock@globnix.org>
- Date: Fri, 2 Jan 2004 22:44:28 +0000
- In-Reply-To: <Pine.LNX.4.44.0312300919020.29088-100000@q7.q7.com>
- Mail-Followup-To: sage-members@usenix.org
- References: <1072799826.7585.24.camel@localhost> <Pine.LNX.4.44.0312300919020.29088-100000@q7.q7.com>
- Sender: owner-sage-members@usenix.org
On 2003-12-30 at 09:23 -0800, Joe Pruett wrote:
> anyone want to share horror/joy stories about:
> snapservers
> performnas
> netstor
>
> specifically as compared to a netapp box. or anyone used lvm snapshots
> heavily to mimic the auto snapshot feature of the netapp?
Which reminds me: whilst FreeBSD 5.x reminds not-quite-stable-enough for
solid production use (ugh), there is some really nice stuff in it. I've
played a little with the new UFS snapshots, enough to verify that they
work on a very lightly loaded system, when not kept around for long.
Has anyone tried these UFS snapshots (mksnap_ffs) in anger yet? Any
information about them at all?
NetApp Filer snapshots rock, but we can't justify NetApp costs for
everything where we want more resilience (customer-facing stuff gets
NetApp with battery-backed write commits, etc, internal stuff often
doesn't). Has anyone tried FreeBSD 5.x with snapshots and RAID 5? Any
particular issues?
More long-range, does anyone know if any current projects are working at
bringing undelete() into production stability, as opposed to a
semi-supported 4.4BSD-Lite feature fading into obsolescence? (_Please_
prove me wrong with that description!)
TIA,
--
"Science is the belief in the ignorance of experts." -- Richard P. Feynman