[2124] in SIPB-AFS-requests
Moved external disk to ronald-ann2
daemon@ATHENA.MIT.EDU (ghudson@MIT.EDU)
Mon Sep 4 03:41:12 1995
From: ghudson@MIT.EDU
Date: Mon, 4 Sep 1995 03:40:38 -0400
To: sipb-afsreq@MIT.EDU
First, the part that affects you: hold off on allocating large sums of
disk space until I report on the success of moving data onto the
Seagate disk on ronald-ann.
The rest:
I've moved the new Seagate 4GB disk from rosebud to ronald-ann2.
(Total outage time was about 25 minutes, with rosebud up after about
ten minutes.) They're now /vicepf and /vicepg on ronald-ann2.
I'm going to start moving the data off of rosebud2 b and c onto
ronald-ann2 f and g, even though I'm not planning to replace the
Micropolis disk on rosebud in the near future. My purpose is to test
whether the vos moves start failing and vos creates become slow like
they did on rosebud. If they do, then we probably have a disk problem
and we shouldn't replace the system board on rosebud, but we should
consider replacing the Micropolis disk on rosebud. If we don't see
move failures, then we almost definitely have a problem with rosebud,
and we should replace rosebud's system board but not the Micropolis
disk.
(Of course, this might be some insidious system-level interaction
which we can't predict, but I'm hoping we're just seeing subtle
hardware failures.)
If the vos moves start failing moving data onto ronald-ann2 f and g,
then I will send back the Seagate swap disk and ask for two 2GB disks
instead. If we continue to have significant trouble expanding the
SIPB AFS cell, I'm going to give up and ask for a Sparc 5 to replace
one of our Decstations. All of this subject to circumstance and
objections.
