shutdown -ry hanging system

[Joe Robinson_2]

Good morning!

I've come across a problem that I am scratching my head on...when I reboot my system (L2000, UX 11.00) using shutdown -ry 0, all logins to remote terminals are ended, but the system will sit in a multiuser mode (runstate 3). When I go to the console and reboot, the command then works normally (I haven't tried rebooting directly from the console recently, so I can't say if the shutdown command works normally there). I don't have any limitations in place (at least not the common ones I'm aware of), but would like to remedy this to make sure the server is stable (I DON'T! like problems during startup/reboot/shutdown!)

Thanks in advance,
Joe

[Helen French]

Check shutdownlog and rc.log.old in /etc directory and look for any errors during last shutdown. Check at what point the system hangs? Any network issues? Do you have any NFS mounts active during shutdown?

[Scot Bean]

The shutdown cmd runs a bunch of /sbin/init.d "stop" scripts. It is possible one of these is hanging.

Does the "reboot -q" cmd also hang?

[Steven E. Protter]

In the shutdown log /etc/shutdown

The last thing that happened before the hang is going to be listed.

Normally, when this happens to us, its because a shutdown script /sbin/init.d/ script is not working right.

A common cause for us is Oracle not coming down clean. We had to write a special program to replace dbshut to get clean Oracle shutdowns.

SEP

[Joe Robinson_2]

There isn't anything in any of my log files to indicate that there are any apps/processes hanging; that being said, I haven't checked to see oracle's status before rebooting from the console. When I run the command at the console, the system behaves normally - I was just wondering if there was any security mods that might make my server act this way (at one time I was using securetty to keep root from logging in except for the console, but no longer am). I have to patch this system this week (tonight, I hope) and will monitor the db status as the shutdown works.

[Helen French]

I remember reading a document about similar type of issue, in which the list_mode parameter was the problem. Check the value of this parameter in /etc/rc.config.d/list_mode file and make sure it's the default - 0.

If the value is 3, that can give an error during the execution of the shutdown script.

just a guess ...

[Joe Robinson_2]

Shiju,

0 is my default; thanks for the suggestion...

Joe

[Todd McDaniel_1]

Do you have any NFS mounts? I know that will hang a shutdown...

I have had to login the console and kill shutdown and execute a reboot to get it down.

[Sanjay_6]

Hi Joe,

If you are trying to reboot the server, have you tried,

shutdown -y -g0 -i6

Or to simply shutdown,

shutdown -y -g0 -i0

Hope this helps.

Regds