TTYs & Printers

[alfred0]

AIX/perts,

i just wanted to copy exactly my "OLD" RS6K machine's defined ttys and printers (both remote and network) - to our "NEW" RS6K machine. are there specific files that needs to be considered?

can somebody give me some idea on how-to deal with it? old RS6K has aix 4.2 and new RS6K has aix 4.3.3. new RS6K will be replacing old in the same network.

HEELLPPPLLZZ!

 

[gileb]

what i did on my system (for printers) was a tar of /var/spool/lpd and a tar of /etc/qconfig.
I restored that on the other system and it works (of course every printer has to be resolved in dns or /etc/hosts)
- i usualy do that for the hacmp systems to put the definitions on the backup nodes.
For the ttys you should try to extract everything from the odm...



hope that helps.

 

[aixqueen]

well for printers

Migrating printers to another machine

Here are the instructions I received from IBM. NOTE: Be advised that IBM
claims they do not support this through IBM.

Migrating Print Queues:
1. copy the /etc/qconfig (make a copy of the original before)
2. Copy the following files:
/var/spool/lpd/pio/@local/custom files
/var/spool/lpd/pio/@local/dev files
/var/spool/lpd/pio/@local/ddi files
3. change the permissions on the copied files to 664 chmod 664 filename
4. redigest the queue's
enq -d
5. run the chvirprt command on each queue
chvirprt -q -d
Note; If you have a lot of queues, the best thing is to create a short script to
do this for you. I grab the queue name and device name by running lsallq -c
and saving it to a file.
6. cycle the printer subsystem
stopsrc -cg spooler
startsrc -g spooler

and you can try doing the restore from mksysb 128 port...for just printers and ttys

http://techsupport.services.ibm.com...mode=9&documents=390605212514786&database=aix

 

[alfred0]

JUST A FEDDBACK....

aixqueen's right re:

====================
Here are the instructions I received from IBM. NOTE: Be advised that IBM claims they do not support this through IBM.
====================

i just did and it messed up my "/var" FS. severe casualties will be your sendmail files.

i would suggest to get all printers' setup/config from the original machine and re-do it. it's better than screwing-up your new system.

fortunately, it works for gileb.

 

[alfred0]

due to what happened, is there a way i can recover original /var FS? will fsck and logform works under service mode?

thnx

 

[aixqueen]

I really am Sorry that it messed up you system. As it said, it works for some and not for others.and IBM does not support it.......the best way is to redo it from scratch. Most users do not want to redo it all from scratch.... We have had a 80/20 rate of success....but sorry you are a casualty.

What exactly is wrong with your var? You should have just done the selected filesystems the /var/spool/lpd/pio filesets not all of var? The sendmail files should have been in other directories...not the /var/spool/lpd/pio? but I may be wrong. The only real fix is to restore the original directories you copied over...if it was all of var...then var...or you may be looking at a preservation install to put it back to the same operating system level...

Yes fsck works in service mode as does logform. But depending on what was done
restore wise, it will determine what the fix it.

 

[Guest_imported]

i got this error when trying to add up queue to a local printer.

enq: (WARNING): 0781-039 Qdaemon appears to be dead.
enq: errno = 2: A file or directory in the path name does not exist.
enq: (WARNING): 0781-162 Cannot awaken qdaemon (request accepted anyway).
enq: errno = 2: A file or directory in the path name does not exist.
enq: (WARNING): 0781-039 Qdaemon appears to be dead.
enq: errno = 2: A file or directory in the path name does not exist.
Added print queue 'lp0'.

and when you try:

root pts/3 $ lpstat
Queue Dev Status Job Files User PP % Blks Cp Rnk
------- ----- --------- --- ------------------ ---------- ---- -- ----- --- ---
lxmk hp@lx UNKNOWN
lp0 lp0 UNKNOWN

and for unknown reason sendmail won't startup. i got no response.

 

[aixqueen]

There appears to be a full filesystem or a corrupted file.

0781-162

Cannot awaken qdaemon
One cause is full /var file system
Another cause is bad /etc/qconfig file
A bad file being printed was one scenario. Cancel, stop and restart the qdaemon.
In other scenarios there were bad files in /var/spool/lpd/qdir.
One fix: simply recycle qdaemon (stopsrc -s qdaemon, startsrc -s qdaemon).
Another fix: clear queues and recycle qdaemon.
Another possible cause: permission problems on /var/spool/lpd/stat
In one case the writesrv was not active. (stopsrc -cg spooler, startsrc -g spooler)

 

[Guest_imported]

thanks aixqueen,

i was able to revert back original /var by restoring it from the backup. so far -so good.