[2461] in SIPB-AFS-requests
hardware?
daemon@ATHENA.MIT.EDU (Jonathon Weiss)
Thu Sep 19 17:51:54 1996
From: Jonathon Weiss <jweiss@MIT.EDU>
To: star-maintainers@MIT.EDU
Date: Thu, 19 Sep 1996 17:51:30 EDT
The maintainers of various SIPB services need to think about what
server hardware should be upgraded this fiscal year (7/96 - 6/97).
This should include CPU's, disks, sbus cards, memory, and other things
of that nature. It might be useful to think about both a minimum of
what must absolutely be replaced this year, and a more liberal
everything we'd like to upgrade this year. Probably we'll end up
somewhere between the two.
Things to keep in mind when deciding what should be upgraded:
1) ASO is planning on upgrading as many of our DEC servers as we have
time to get to this fiscal year
2) The oldest full height disks will be 5 years old next summer, I
believe that is when their warantee runs out. With the exception on
the dialups, I believe ASO has removed these from all of our servers
already, and are working on the half-height 1G drives.
3) are there any servers that are having load related problems that
can be solved with more/bigger disks/memory or a better CPU.
4) we should assume RTFM services on penguin since we've had it for a
couple of years now
I took the liberty of giving some preliminary guesses to the people
who control hardware renewal budget, since they wanted to have some
idea what was going on, I had to provide the same thing for ASO, and
the SIPB and ASO server renewal money comes from the same place.
However, I would like to get amore definate estimate fom the
maintainers of the various services.
I guess what I expect to see is each set of server-maintainers decide
what their service needs, and then have maintainers of all services
get together to come up with a prioritized list of what SIPB
needs/would like. This has been done in the past, but it has been a
couple of years, I think. However, if someone has a better way to do
this, I'm not particularly attached to my method.
Jonathon
