[3066] in SIPB-AFS-requests
Re: FY98 budgeting and AFS
daemon@ATHENA.MIT.EDU (Greg Hudson)
Wed Jul 15 22:58:07 1998
To: John Hawkinson <jhawk@MIT.EDU>
Cc: sipb-afsreq@MIT.EDU
In-Reply-To: Your message of "Wed, 15 Jul 1998 22:25:53 EDT."
<199807160225.WAA03286@zorkmid.mit.edu>
Date: Wed, 15 Jul 1998 22:57:50 EDT
From: Greg Hudson <ghudson@MIT.EDU>
> I'd be a little concerned that the RAID may introduce unfortunate
> software dependancies ("No, you can't run Solaris 2.7, you have to
> wait for the RAID vendor to upgrade their software") that might
> prove suboptimal.
I believe (someone correct me if I'm wrong) that this unit just acts
like a big SCSI disk, so it's not more likely to cause this problem
than a disk.
> I dislike the concept of a single monolithic disk attached to a
> single server. I'd be significantly more comfortable with, say,
> multiple 20gb RAIDs.
The two problems with smaller RAID units are (a) that they're not much
cheaper than larger RAID units and (b) that none of them have been
qualified by ASO.
> Restoring a 72GB unit is a Big Production.
This is a big objection, if we think it's likely that a whole RAID
unit will go and take all its data with it more often than, say, an
AFS server with four 9GB disks. I'm not sure how true that is, but we
probably have to assume for now that the risk of that happening is
non-minimal.
Given the reservations from jhawk and mhpower, I will back off on this
plan, and instead propose buying one or two new servers in FY98 and
replacing some number of our older disks. But I will wait on that
until we have gotten farther on deploying our existing new hardware,
or until someone asks us to draw up an FY98 budget.
