Unable to boot - console error messages

[steven Burgess_2]

Hi all

I am presuming this server already had issues -

whilst performing useradd -m

received error message

mkdir : unable to create /home/

System then froze

Checked the console and nothing - Hard rebooted machine - returned with errors

INIT:command is responding to rapidly

will try again in 5 mins

id:cons /usr/sbin/getty console console
id:vue /usr/bin/vue/vuerc

I have checked and found a few links with regard to these errors -

1st - Check /etc/rc.config.d and search for any script names with unusual characters - None found

2nd Ensure entry in /etc/inittab is correct for the console and vue

If I was to comment out the entries , would I be able to log onto the server without getty?

I am unable to use commands such as more and vi ,If I attempt to mount -a , I get message regarding h/ware paths - vg01 - device not found at address

Have I got a possible disk issue here as well

Any comments would be great

Apart from - Go Home and wait for the england match to kick off -

Regards

Steve

[Mike Hassell]

Steve,

Sounds like a hardware problem, either a disk or a controller. Do you have STM installed or EMS? What errors do you have in your syslog? I would start with the following:

ioscan -funC disk

Look to see if you've got the disks that are normally there. Then try running xstm or stm depending if you have Xwindows support at your term and run hardware diags on your disks and controllers, it should give you an idea real quick.

Be sure to check your cabling as well to ensure you've got a good connection with all your disks.

I assume this is a test or development machine? or at least I hope :-)

Good luck!

-Mike

[steven Burgess_2]

Hi Mike

Yes - a test machine -

Have

cat /etc/inittab

The entry for cons doesn't have -h - will this cause an issue ?

Have checked ioscan - all disks are claimed

will check cstm

Thanks - back shortly

Steve

[T G Manikandan]

Steven,
First check your /var/adm/syslog/syslog.log for any error messages.
Just check there whether you have any power failed disks.
Second check whether that file system is a NFS mounted one and check whether the remote machine is up and running and you could reach it.


Thanks
G Manikandan

[S.K. Chan]

The "INIT: command respawning too rapidly" error can be a nuisance because it's not easy to pin point a straight forward "root cause". I can think of a few things/reasons, which you can check them out yourself ..
1) Of course the obvious one .. you got a corrupted or bad /etc/inittab file. You can try to copy the default "inittab" from /usr/newconfig/etc and see if it helps. Make a copy of your existing "inittab" first.
/usr/newconfig/etc/inittab.
2) Startup script and its configuration file having problem. Typically /etc/rc.config.d; /sbin/init.d and all the other /sbin/rc[1-4].d. Are there any changes made that you're not aware of ? Check for the timestamp for recently modified files, check for permissions, "unlink" scripts in /sbin/rc[1-4].d.
3) /etc/fstab file corrupted. Yes this can be one of the reason too.
Do you see any other errors in "dmesg" or /var/adm/syslog/syslog.log that could possibly point to a disk failure (the other about hrw path that you're getting)?

3hrs30mins to kickoff ..

[Mike Hassell]

Steve,

I don't know about the -h on cons, but a typical entry in /etc/inittab should look like this:

cons:123456:respawn:/usr/sbin/getty console console

Thus getty will be running on the console in every runlevel.

Take a look at what processes are running as init shows that some command is responding too rapidly. Do you see any abnormailiy in a:

ps -ef

Also check for any typos in /etc/inittab as that could be throwing things off.

The lack of having access to commands like more and vi puzzles me though. Also are you sure you need to be running vue from inittab? What OS level are we talking about as well? Sounds like an older box that has some issues that need to be worked out. Be sure to look at /var/adm/syslog/syslog.log as it may shed some light on this whole thing.

Hope you get it worked out.

-Mike

[steven Burgess_2]

Hi all

Box is an old D class running 10.20

There are files in /var/adm/crash with a timestamp of when the box died

I can't look at the syslog bacause it's not there -

I can't copy /usr/newconfig/etc/inittab - not there either

dmesg - shows no errors

Unable to use cstm - not found

Can only use intrinsic commands

I think I may leave this for my seniors on days as they will know if there are outstanding issues with this box -

I have one hour to go before I leave

Further suggestions welcome

2.5 hours till kick off

Come on England !!

Steve

[Michael Tully]

Hi Steve,

If you have a support contract for the server, get the crash dump analysed. If not maybe try your hand at using the gdd/adb commands.

Another suggestion, seeing it is a test system why worry, patch it, reinstall the OS or even upgrade if you can. The error is coming from /etc/inittab file. Do you have any weird printer setups, serial printers ?

Go Senegal.... just kidding!

[steven Burgess_2]

Hi Micheal

Last one for Today - no peripherals at all attached to the server

I did , out of curiosity attempt to activate vg01 - but got errors related to a disk on there

I think a fresh install or boot from ignite may be the best option

Thanks all

Steve

[Santosh Nair_1 · Accepted solution]

Sounds like you have /var on /dev/vg01. And if you're saying that you can't activate vg01 because of some disk problems, then yes, you will have issues. Your best bet (after replacing the failed drive) is to recreate the OS from a bootable tape (if you have done a make_tape_recovery). Otherwise you probably will have to re-install the OS.

-Santosh

[John Bolene]

Yup, sounds like disk issues.

I have a similar type problem with one of my test machines and am just waiting for the disk to finally die instead of being intermittent.

The machine will work for between an hour and several months but eventually gets a disk hang and has to be powered off and back on. Then it works again. TOC will not even do anything. No dump is ever made. When hung, the box returns a ping but nothing else.

Restore that machine with an ignite image if you have one.

[John Carr_2]

Steve

if you can not run xstm or stm it is because you do not have the Online Diagnostic Environment installed this should be loaded on all systems as a matter of course you will find it on the 10.20 support CD if you dont have the CD its available as a download from the HP download site.

John.

[Martin Johnson]

Boot the system using the maintenance CD and run fsck on the disk. This will check for file system errors.

HTH
Marty