[2201] in SIPB-AFS-requests
More problems with the Seagate 4GB disk
daemon@ATHENA.MIT.EDU (ghudson@MIT.EDU)
Tue Jan 2 02:01:54 1996
From: ghudson@MIT.EDU
Date: Tue, 2 Jan 1996 02:01:40 -0500
To: sipb-afsreq@MIT.EDU
The outage I just scheduled is to reorder the SCSI bus on ronald-ann2.
ronald-ann2:/vicepg is being slow in the same way the disk was on
rosebud2:
packet-drop% time vos create ronald-ann b test.vol -c sipb
Volume 536877307 created on partition /vicepb of ronald-ann
0.23u 0.23s 0:02.61 17.6%
packet-drop% time vos remove ronald-ann b test.vol -c sipb
Volume 536877307 on partition /vicepb server ronald-ann deleted
0.24u 0.21s 0:02.28 19.7%
packet-drop% time vos create ronald-ann g test.vol -c sipb
Volume 536877328 created on partition /vicepg of ronald-ann
0.23u 0.25s 0:23.37 2.0%
packet-drop% time vos remove ronald-ann g test.vol -c sipb
Volume 536877328 on partition /vicepg server ronald-ann deleted
0.23u 0.21s 0:01.98 22.2%
As before, we also had a vos move fail:
packet-drop% vos move linux.slackware.nb rosebud2 d ronald-ann2 g -c sipb -verbose
Starting transaction on source volume 536873666 ... done
Cloning source volume 536873666 ... done
Ending the transaction on the source volume 536873666 ... done
Starting transaction on the cloned volume 536877305 ... done
Creating the destination volume 536873666 ... done
Dumping from clone 536877305 on source to volume 536873666 on destination ...Failed to move data for the volume 536873666
vos move: operation interrupted, cleanup in progress...
clear transaction contexts
access VLDB
move incomplete - attempt cleanup of target partition - no guarantee
cleanup complete - user verify desired result
(There may be a turd left around from this; the volume is still
accessible, fortunately.)
You will recall that after replacing the disk twice (with Seagate
disks) and after significant tests on rosebud2, we were unable to
determine the nature of the problem. We had rosebud2's system board
replaced, but the problematic disk wound up on ronal-ann2 d.
The theory is that if reordering the SCSI bus solves the problem, it's
a SCSI bus problem and we can have the system board on ronald-ann2
replaced. If reordering the SCSI bus doesn't help, then the only
course of action I can think of is to replace the Seagate disk with a
Micropolis disk. (I guess we could try replacing the Maxine with a
Sparc 5 or Sparc 20.)
