Terminal used to restart NBU locks up......

[Joe_Despres]

Has anyone seen this:

When I login to a NBU server and restart the processes...

My terminal locks up when I attempt to logout....

Is there a fix or a work around for this?

Thanks.....

Joe Despres

[lu]

You can try to run something like 'nohup /etc/init.d/netbackup restart'

[Karthikeyan_Sun]

What is your Event Viewer Saying?

[J_H_Is_gone]

what is master (win/unix)
are you telneting or sshing into the server? (explain terminal)
what command are you using to start the processes
do you get the command line prompt back once you do?

[Joe_Despres]

Master = Solaris 9
Media server(s) = Solaris 9

Terminal locks up using the following methods:
a::    ssh into servers using  xterm from within cygwin...
b::    ssh into servers using rxvt from within cygwin....

sudo /usr/openv/netbackup/bin/goodies/netbackup stop
sudo /usr/openv/netbackup/bin/goodies/netbackup  start

I do get the prompt backup.....  Problem starts when I exit out of the terminal...
NetBackup stays up....

nohup is a good idea..................

Thanks...

Joe Despres

[lu]

Yes use nohup with the start script: nohup sudo /usr/openv/netbackup/bin/goodies/netbackup start

[J_H_Is_gone]

nohup is only going to allow a job to keep running once you log out.
as the netbackup start is a job that shold come to compleation... and from what he says it does.. as he gets a prompt back.. I don't think nohup will help.

.. now my question is how do you know your session hangs when you exit?

[Stumpr2]

Netbackup stops and starts and returns you to a shell?
Then, this appears to be an xterm related problem.

[Joe_Despres]

This doesn't affect NetBackup.......

It 's just I would like to automate the stopping and starting of NetBackup.....

I can kill NetBackup remotely....  It's the restarting of NetBackup where I hit a brick wall....

If I can't do it manually...  then automating it won't  work either....  I have tried....

Both xterm & rxvt terminals exibit this problem....  As well as putty......

Joe Despres

[J_H_Is_gone]

lets look at the steps

you telnet/ssh to your master server
you get a login prompt
you log in with your user name and password
you get a command prompt
you use sudo to run netbackup start  or stop
you get the command prompt back (this means that it worked and it is not a netbackup issue)
you type in exit

.......

so what is happening that you say it is hanging up?
you have a prompt, you type exit, ........... and what?

[Will_Restore]

nohup 'sudo /usr/openv/netbackup/bin/goodies/netbackup start' < /dev/null &

[Joe_Despres]

It justs hangs..... 

Should give me the prompt backup which is on my PC...

Or it should go away....

Here's how I do it:

Login to the master:
rxvt +sb -sl 1000 -T "master" -e ssh -t master &

Login to media1:
rxvt +sb -sl 1000 -T "media1" -e ssh -t cscnb01 "ssh -t -C media1 \"$@\"" &

Login to media2:
rxvt +sb -sl 1000 -T "media2" -e ssh -t cscnb01 "ssh -t -C media2 \"$@\"" &

Login to media3:
rxvt +sb -sl 1000 -T "media3" -e ssh -t cscnb01 "ssh -t -C media3 \"$@\"" &

and the list goes on for 15 media servers...

I have all this in a script....  I kick it off and poof..  I have a terminal open on all my NBU servers...

Joe Despres

[Stumpr2]

try it again but instead of logging out run the commands to see what you may be running

$ ps -ef  | grep <loginID>

$ jobs

$exit  <-- if it hangs then leave it alone and do a second login <different loginID> and see what processes are still running with the first loginID

 

[lu]

I just had the same problem on a media server and the culprit is... vmd With "lsof", you can see that it's not properly detached from the terminal: vmd 20079 root 0u VCHR 24,1 0 /devices/pseudo/pts@0:1->ttcompat->ldterm->ptem->pts Hope to see this bug fixed in a next release....

[lu]

Even with a nohup, vmd still opens /devices/pseudo/pts@0:1 ... Very annoying...

[lu]

This WA works for me ! : /etc/init.d/netbackup start </dev/null

[Joe_Despres]

I tried this with some success:

 sudo /usr/openv/netbackup/bin/goodies/netbackup start  >/dev/null  2>&1

[Omar_Villa]

what happens if you use the bp.kill_all script?

[lu]

No not >/dev/null, but </dev/null /etc/init.d/netbackup start </dev/null

[lu]

The problem has nothing to do with stopping daemons, only with starting them...