Hi,
Original message:
We have had a continuing problem with an AlphaStation 200 4/233, with
DU4.0B, where the thing just stops in the middle of the night. 3:10am
today, 1:20am Sunday, and once at 9:30am. Not while anyone is actively
working on the machine. It leaves no tell tail messages in the error log
as to what is going on. It would appear that it might be related to the
internal system disk, I only say that because it generally will not boot
until the machine has been switched off, as it can't find the boot file
during the automatic boot after a halt.
To find out what time the machine was falling over I just wrote to a log
file, at 10 minute intervals. I've checked the syslogd.conf file, and it
appears to be normal, probably more things could be switched on, but if
the machine knows its about to go down, it sure does think its critical
enough to record at the time.
DEC have so far replaced the (a) mother board, (b) power supply, (c) the
halt switch, to no avail. The machine doesn't have night activity except
for a backups, which have been finishing around 12:00pm. Occasionally
the machine has crashed during a backup, but generally the machine
prefers to crash when it has nothing else better to do.
This is the first time I've seen this, and was wondering what part of
the machine may be causing the thing to shutdown, without so much as a
wimper
Solution:
It turned out that the boot disk was comming up with scsi errors in both
/ and /usr, these were detected by switching the console to serial and
attaching a laptop to COM1. Why these errors generally occured during
the night and why they didn't appear in the syslogs I'm not sure. The
replacement boot disk has been working fine for 2 weeks, so maybe the
case is closed. I might add that a new installation of 4.0b was used
rather than copy the one from the faulty disk.
Dennis
email: mcdonell_at_auslig.gov.au
Received on Tue Feb 17 1998 - 00:04:11 NZDT