L 2000 hangs on startup

[Michael O'brien_1]

Hi,

I have an L2000 server which hangs just after if reaches the autoparms message. If I boot the machine into single user mode and then take it throught init1, init2 and run the scripts in init3 on at a time. Then the server has no problems. If I try and let the server boot normally the if hangs just after the autoparms message. If will just sit there and not continue.

If I press "control \ c" then enter it will then boot.

What processes are happening after the autoparms message and before the first entry in rc1.d?

If there's a problem with a script in run level 3 whay doesn't it boot until if reaches the scripts with the problem?

Why will it boot if the take through the run levels maually?

If anyone can give me some pointers of what might be causing this problem I would much appricate it.

Many Thanks
Michael

[Steve Steel]

Hi

Check if DHCP is enabled in /etc/rc.config.d/netconf but not installed.

That could do it.

What is the exact message


Steve Steel

[Michael O'brien_1]

I've check the /etc/rc.config.d/netconf,

The DHCP is set to 0

[Peter Kloetgen]

Hi Michael,

it is very easy to find out exactly, which scripts are run when entering a new runlevel:

in the directory for scripts to be run when entering runlevel 3 you do a listing and all more you have to do is to look at the names of scripts:
S100name_of_service
S110name_of_another_service

and so on....

The number after the S is the passnumber of the script! The script with the lowest number is executed first, after it the others, so you can exactly see which script is run at which moment. ( The same procedure is also taken for the kill- scripts of course!)

Allways stay on the bright side of life!

Peter

[Michael O'brien_1]

What I need to find out is what is happening after the initial boot phase and prior to the first run level starting. After the auto_parms message and prior entering the rc1.d and running the crashconf script.

Thanks

[Peter Kloetgen]

Hi Michael,

when the initial sequence is over ( the copy of the kernel is in RAM ) the swapper daemon is started ( PID 0 ) and after that control is taken by init. That command first reads the /etc/inittab to find out the initdefault. ( Normally 3 ) Take care you have a good entry in that line! ( Key word is initdefault )After this, init reads the following lines with the possibilities:

first field : name entry
second field : defines, in which runlevel the commands should be done. No entry here means in every runlevel. One entry means in only one runlevel, several entrys means in the defined runlevels.

You should check if all entries are correct. If the answer is yes, then find out if the start- and stop- scripts have the corret pass- number ( see my first mail ). Where any changes made, and if the answer is yes, where have the changes been made. There exist dependencies!

Allways stay on the bright side of life!

Peter

[melvyn burnard]

Hmm, what happens if you move /sbin/auto_parms out of the way?
e.g. cp /sbin/auto_parms /sbin/auto_parms.suspect
and then clear out auto_parms so it's file os 0 size? or even just remove it?

I suspect there may be an error in this file, as there are some settings in there that could mistakenly be set to try to use DHCP etc.

[Ravi_8]

Hi,

This shows that auto_parms file has been corrupted. Get into single user mode
# cp /sbin/auto_parms /sbin/auto_parms.old
#set_parms initial
then reboot the machine would solve the problem