Memory fault(coredump)

[zhang wenhua]

I have a hp 9000 ,HP-UX B.11.00 A 9000/785 2006439932 two-user license.On which run ITO and NNM
It response with Memory fault (coredump) when I run "last" command.
The customer have deleted all files below /var/opt/OV/share/tmp. I can't start ITO now.
Who can give me some suggestion?
Bests Regards

[Olav Baadsvik]

Hi,

It could be that your wtmp file is corrupt
or not existing.
/etc/wtmp is a link to /var/adm/wtmp.

/etc/wtmp has the following owner/permissions:

rwxrwxrwx root sys

/var/adm/wtmp has this:
-rw-rw-r-- adm adm

Regards
Olav

[Deepak Extross]

Just deleting /var/opt/OV/share/tmp should not cause a coredump when you run 'last'. I suspect some other directories have been messed up as well.
The first step would be to find out exactly what all has been mucked up.
Looks like you may have to reinstall ITO, but first check teh extent of the damage.

[Peter Kloetgen]

Hi Zhang,

you seem to have a corrupted utmp or wtmp file, these are used by commands like last or who to get their information. You mentioned, that you are using ITO/OVO. Simply call your support team, they created a tool which can repair these files when they are corrupted. This should fix your problem.

Also you can have a look to the manpages for utmp and wtmp, there is a fix- command described on them, perhaps this will help you....

Allways stay on the bright side of life!

Peter

[Steve Steel]

Hi


these are all useful pieces of advice.

If they do not solve it.

1)What is output of
file core
on the memory dump file.

2)Does lastb work

3)Are there any messages in /var/adm/syslog/syslog.log for this time.

Steve Steel

[Wodisch]

Ni Hao,

what is the output of your "ovstart -v" and if the NNM part is starting, the output of "opcsv -start" (or "opcsv -status" else)?

Could it be your "user" has destroyed some libraries, too?

Zai jien,
Wodisch

[Lisa Mauer]

I have the same thing happening on one of my servers, I can do "lastb", but "last" causes a Memory fault(coredump.

The output from the dump is:
core:core file from 'last' - received SIGSEGV

There aren't any messages in the syslog.log file...

I checked the permissions and they are the same as Olav stated in his post:
/etc/wtmp has the following owner/permissions:

rwxrwxrwx root sys

/var/adm/wtmp has this:
-rw-rw-r-- adm adm

Anything else?
Thanks!

[David Burgess]

I've got :-

lrw-rw-r-T 1 root sys 13 Jan 18 16:40 /etc/wtmp -> /var/adm/
wtmp

-rw-r--r-- 1 root root 900 Apr 30 14:34 /etc/utmp

In var/adm I've got :-

-rw-rw-r-- 1 adm adm 57120 Apr 30 14:34 wtmp
-rw-r--r-- 1 root sys 280 Apr 30 11:26 wtmpx

last and lastb work fine.

HTH,

Dave.

[David Burgess]

Can you restore the missing files?

Regards,

Dave.

[Lisa Mauer]

Fix for "last" or "lastb" memory core fault...

"last" command gets its information from /var/adm/wtmp file where as "lastb" gets it from /var/adm/btmp.

In my case I could not run "last" but could run "lastb" so I ran the following command:

"cat /dev/null > /var/adm/wtmp"

and now "last" works perfectly.

Seems as though my wtmp file was corrupt.

Hope this can help you.