Thanks to Alan from dec.com for this:
The errno value 5 is for I/O errors. The 8,1030 is the
major and minor device number of the device with the
problem. The major number 8 is typically the SCSI
disk driver. Minor number 1030 is probably rz1g.
My guess is that you have UFS on the particular disk,
which allows failed writes to sit in the buffer cache
until they succeed. This has the advantage of being
very tolerant of transient errors. It has the disadvantage
of allowing the file system to become very inconsistent
when a disk fails in a way that will prevent it from ever
getting better.
and Jim Belonis for:
Dead controller, bad SCSI cable, all the usual suspects.
The machine failed during the night, I booted to single user mode, ran fsck
on both drives and got no errors. I rebooted it and it is working
perfectly. I think that I have a controller going bad, I won't know for
sure until I analyze the logs or it fails again.
-----Original Message-----
From: Haner, Kenneth C. (OAO- HOU)
Sent: Monday, March 12, 2001 4:32 PM
To: tru64-unix-managers_at_ornl.gov
Subject: No login
I've got a Alphaserver 1000 running Digital Unix V4.0D that stopped
responding to telnet requests. I went to the console to login and cannot
get a prompt. The screen is scrolling error messages such as.
Deferring I/O (errno 5) for block 0x40,0x40 on device 8,1030
Deferring I/O (errno 5) for block <these numbers are changing> on device
8,1030
occasionally I get the same message except (errno -1). The machine is still
responding to http requests...it has two drives in it. One has a solid
green light, the other is blinking. I'm guessing the blinking is the drive
trying to read...ie. dead hard drive. Before I buy a new hard drive does
anyone know what else might be causing this?
Kenneth Haner
khaner_at_houston.oao.com
Received on Tue Mar 13 2001 - 14:45:27 NZDT