From: Mike Knight, IBM GS Central, (314)234-5096, KNIGHTM at ISSCVM
Inet: KNIGHTM AT VNET.IBM DOT COM
Subject: Scheduler has been stopped
On Tue, 20 May 1997 13:46:18 +0200, Peter Duempert added:
>4. Comments:
> 1. In the above log I did a restart of the dsmc-client at 12:58 due to
> "Scheduler has been stopped"
> on the day before
> 2. At 17:15 the same day again the:=20
> "Scheduler has been stopped"
> 3. A "core"-file has been produced as:
>-r-------- 1 root sys 418652 Mai 19 17:15 core
> 4. running "xdb" with this core, delivers no useable info ( dsmc is
> "stripped" ? )
> 5. No further evidence in any of the system-log-files.
> 6. If I run this scheduled backup manually, all works well.
This appears to be the same problem we have with some HP clients. The
problem comes and goes and has happened on different ADSM client and
server patch levels. At least one of the failing clients is HP-UX 9.05,
I haven't checked the others. We will have it for a week and then
backups will work for a month. We haven't found a cause.
In our case, APAR IC16534 described the symptoms and we decided to wait
for a fix instead of duplicating the effort. But the APAR was cancelled,
so I plan to open a new problem and see what happens. For us, and APAR
IC16534, the last thing in a client trace was
fsinfo.c (xxxx): Second mntctl failed, errno = 2
and the "Scheduler has been stopped" message in dsmsched.log.
We also get the same failure with a manually started incremental backup,
but have found that both the scheduled and manual backup will run if
started shortly after another ADSM backup has failed. The first one
fails and the second works with no changes. I threw together a short
script to restart the client dsmc schedule when it stops. This seems to
get the backup to run near the end of the window when the backup starts a
minute after the client schedule process is restarted.
Mike (Just another user, not ADSM support) Knight
|
