[1517] in SIPB-AFS-requests
Re: machine room orginization
daemon@ATHENA.MIT.EDU (Jonathon Weiss)
Mon Oct 3 15:25:27 1994
To: Greg Hudson <ghudson@MIT.EDU>
Cc: Jonathon Weiss <jweiss@MIT.EDU>, star-maintainers@MIT.EDU
In-Reply-To: Your message of "Sat, 24 Sep 1994 02:23:09 EDT."
<9409240623.AA05962@glacier.mit.edu>
Date: Mon, 03 Oct 1994 15:24:52 EDT
From: Jonathon Weiss <jweiss@MIT.EDU>
>> I don't see any point in it right now; it's already pretty easy to
>> trace the SS20 cables. Our section of the machine room is a complete
>> mess in general, but we don't have much hope of cleaning it up until
>> we get rid of the DS3100, the DS5000/200, and the RS6000/520. At that
>> point, it's questionable whether we should cause an outage for all of
>> our services just so we can fiddle with cables. We can, however, make
>> an attempt to rearrange the machines without cutting off power.
I'm sorry I took so long to reply...
The main reason I am suggesting we start this now is so that we can
re-cable each machine as necessary, while it is not in production.
IMHO the SS20 cables didn't look that neat when I looked at them, they
were also plugged into the wall outlet, rather than circuits under the
floor.
I was thinking something along these lines:
1) spread the SS20's out, probably taking one of the small carts for a
little extra real estate (I would guess RTFM's is easiest to steal,
just move it onto the stack on it's left for now.
1a) rewire the ss20's more neatly, and probably moving them to outlets
under the floor. Turn the ss20's around (at least the CPU's for now)
I'd like to end up removing the table the AFS servers are currently
on, so that we are back to our L-shaped server area eventually.
2) As we get beacon and rtfm migrated to SS20's we can clean up the
table that has most of the decstations now.
3) Finish by moving ronald-ann onto the decstation table, and removing
the current AFS table.
This is a kinda sketchy outline, but I have spent some time thinking
about it.
Comments?
Jonathon
