[3064] in SIPB-AFS-requests
Re: FY98 budgeting and AFS
daemon@ATHENA.MIT.EDU (Karl Ramm)
Wed Jul 15 20:06:29 1998
To: mhpower@MIT.EDU
Cc: ghudson@MIT.EDU, sipb-machine-room@MIT.EDU, sipb-afsreq@MIT.EDU
From: Karl Ramm <kcr@MIT.EDU>
Date: 15 Jul 1998 20:06:24 -0400
In-Reply-To: mhpower@MIT.EDU's message of Wed, 15 Jul 1998 17:55:12 -0400
mhpower@MIT.EDU writes:
> >think it would make our cell easier to manage if we started to
> >transition to these units instead of maintaining a plethora of small,
> >differently-sized disks.
>
> I think it might make the cell easier to manage by people who
> already have a great deal of familiarity with the details of the Box
> Hill RAID implementation, perhaps due to working with the
> Box Hill RAID implementation in other afs cells.
It's not rocket science. I'm shocked, nay even dismayed by the fact that
you think that the non-ASO sipb-afs members are a bunch of idiots that are
incapable of dealing with a nice, friendly, menu-based system, or even the
perl libraries that box hill provides. I would submit that if they were as
stupid as you seem to imply that they are, then they probably shouldn't be
running an AFS cell.
Anyway.
Configuration of the Box Hill units is done through this friendly menu
interface that I have seriously played with, that perhaps mwhitson can
comment on. This needs to be done once, barring the catastrophic failure
that I will address in a moment. Now, my understanding is that further
maintenance occurs through the process of noticing that a disk has failed
and replacing it in a reasonably timely manner. Two (of the wrong) disks
failing in the same redundancy group could cause one of those catastrophic
failures that we're worrying about.
However...
Assuming that we add 72 (or more) GB to the SIPB cell, we're going to have
to do AFS backups differently when the backups goes to two or three tapes.
We'd probably want to purchase a DLT stacker (and put it in the machine
room), and at that point, restoring from several tapes quickly and
efficiently is no harder than restoring from one tape, which is our current
situation.
kcr
