On #2, why a scheduled cmd to run a script to run a backup, just run
the backup...
and the continued attempt might not be associated with the backup, it
could be the scheduler trying to establish a session to communicate
the offset of the next scheduled event...
last tid bit... In a guess-duh-mation that the node admins probably
wouldn't have the forsight to set access to another node and not
knowing just what miracle they are going to want me to perform next I
attempted to run a scheduled command that did a DSMC SET ACCESS BACKUP
"*" servernode root -PASS=blahpass (most of the client nodes I don't
even have a plain id on but I do know all the client pswds) Well the
scheduled event FAILED and for some reason the client started in with
looping connections like you have mentioned. I had to have the
scheduler (dsmc sched) killed on that node to get it to stop.
On #3 check sched log... I know it reall only says FAILED in there
also but if your client system has anything like SMF records (ie MVS)
I would look into your serialization options and activity against that
dataset at that time... then maybe an inspect against that data set's
disk area... I have yet to see adsm fail to backup a dataset without
1) it being trash to begin with (I/O errors on the device) OR
2) serialization options saying DON'T DO IT which yields a "failed"
______________________________ Reply Separator _________________________________
Subject: Expire & comm's queries
Author: ADSM-L (ADSM-L AT VM.MARIST DOT EDU) at unix,mime
Date: 2/6/97 8:34 AM
Hi everybody,
First of all I would like to thank all those who have
provided me with some solutions to my earlier questions, but now I have
some more.
Setup AIX 4.1.4/5, Server 2.1.0.11, Client 2.1.0.1
1) I have had the following situation occur where I've been running
incremental backups on an SP2 node. When testing I excluded all
filesystems with the exception of my HOME directory, I then ran an
incremental backup which promptly started to expire files that existed
on the node and had previously only been backed up once, these files
were then also marked as inactive. My query is that, if a filesystem has
been excluded from a backup, then shouldn't all the files within that
filesystem remain active ?
2) When I run a scheduled command to run a script which in turn runs a
selective backup, I keep getting the following message :-
ANR8213W Session open with xxx.xx.x.x timed out.
ANR2716E Schedule prompter was not able to contact client XXXX using
type 1 (xxx.xx.x.x 1501)
This message appears roughly every 2 Mins, although the backup is
running OK. I have checked that the tcpport, commmethod, buffsize &
windowsize are the same on the server & client, there also isn't a great
deal of network traffic when the backup is run and this error doesn't
appear when I run an incremental backup. Can anybody explain why I'm
getting this ?
3) Can anybody tell me if it's possible to find out why a client was
unable to send a file to the server, as this caused my backup to fail,
although it backed up other files after the failed one.
TIA
Darren
|
