LOCK the screen or LOG OFF?

Question

Hello All,
OS: SLES 9
What is the difference between LOCK the screen or LOG OFF?
From the server I run command (rwserver.sh server=repserver90 batch=yes &) to run my report server up or running.
Pl note: If its down I am not able to run my reports.
After that I have lock the screen.
Now I am able to use any client (windows machine) to run reports from my Oracle AS url.
I wonder if I LOG OFF from the server,
is my report server still running?
Regards,
DN

Answer 1

What is the difference between LOCK the screen and LOG OFF from the server?
Well, lock screen keeps all the processes running,
whereas log off from the session, would kill most processes
(not the oracle server, but most others that are still associated
with the X session).Something I want to add to that: Locking the screen also prevents other users from using that terminal session, but it depends on where you're locking the screen. If you are not running an X session and you lock the screen via flock or some similar console locking mechanism then users can't use that terminal session, but can use other terminal sessions on the same machine.
If you are running an X session and you lock the terminal via xlock or xscreensaver, you prevent users from using that computer station entirely. Also, if you're running an X session, utilities like flock become quite meaningless because they don't stop the user from opening a new terminal window.
From the server I ran command to run my report server up.
rwserver.sh server=repserver90 batch=yes &
I think if you use the command with nohup instead,
it would detach from the controlling terminal, so it
wouldn't matter whether you only lock the screen,
or log out completely.The ampersand "&" at the end of the command line throws the process into the background. Most of the time this is enough to dissociate the process from the controlling terminal enough that logging out will not kill the process, so logging out after starting the server that way should leave the server running.
Nohup's behavior changes with the environment you're using. In tcsh, it simply causes the process to ignore hangups, though it can itself be overridden by the process. In Bash, nohup will will also change the scheduling priority of the process and redirect output and error messages to "nohup.out", either in the local directory or in the $HOME directory (and if it can't write to nohup.out for some reason, it refuses to run the process).
Another method is to use the GNU "screen" command.But if you use screen, make sure that the screen command is ABSOLUTELY NOT setuid root. I know most sane distributions nowadays don't install it with setuid root anymore, but paranoia never hurt the admin.