I believe I have found my problem. But before I unfold this story, the
quick on DIAGMON.
I ran DIAGINIT.DIAG.SYS and got DIAGMON already running (paraphrasing
here)
So I ran DIAGMON.DIAG.SYS and the prompt came right back. However,
glance no longer showed
diagmon in the list -- and my tape process was finally getting time.
I thought I would check logtool/memrpt as suggested. So I first listed
LOG?@ and saw many. I checked
the create dates and saw that current log=114 and existing was 115+.
With previous create dates!
I purge all LOG#@ files and NMLG@ files. SHOWLOG said logging suspended.
Did RESUMELOG and
things look better now.
I think I messed myself up with my tape job recovering files which
included LOG#@ files. Last FRI I think
the log rollover ran into one of my existing logfiles. Just a guess.
Now the story with DISCUTIL
After my FSCHECK last week (2weeks?) and a bad sector/block, I though I
would start running FSCHECK
on all our systems. So over the holiday, from home, while jobs where
running, I did FSCHECK CHECKALL
from a job. One system reported a bad sector, and again (same sector),
etc. Looked like a loop. So I
aborted the job (attempted). It wouldn't go. I got other jobs off. Then
I ran as session fscheck and got
locked up. Disconnected tcp connection and tried to logon again. !System
had crashed!
Next day system would not boot -- just after mount volumes it died. HPRC
said sounds like bad disk, need
to replace it. It was LDEV 1. I had no backup of ldev 1,2. While
waiting for HP to show, I ran DISCUTIL
(one of the ISL/ISP utilities). It allows you to store files specifying
a fileset, ldev# or volumeset.
I used @[log in to unmask]@ volumeset MPEXL_SYSTEM.... and got most of ldev 1 when it
hit the bad sector -- guess it
was in the directory area. So I stored @[log in to unmask]@ ldev=2, and again @[log in to unmask]@
ldev=1.
After the disk was replaced, OS Installed, Configs rebuilt (no CSLT),
other user volumes mounted, and
system up, I started the recovery: VOLUTIL: recover ;KEEP.
RECOVER has no options other than KEEP or NOKEEP. So it attempts to
recover/restore all files. Hence
my LOG#@ files.
My favorite error in recovery is NON-EXISTENT GROUP. It does not build
the groups for you. So I have to
run this twice -- 1st to get groups/accts stored, then again after I
created the groups/accts. Spoolfiles
do not recover (some error with file label).
I am currently running my 2nd tape (ldev 1) via job.
Files recovered appear alright. Account/Group caps and access right need
to be fixed.
*********************
* Keven Miller KC7LYD
* [log in to unmask] (Salt Lake City, Utah, USA)
* ICQ 5721825
*********************
|
