inetd daemon getting killed

[Sanjiv Sharma_1]

Hi,

I have 2 standalone K460 with HP-UX 10.20.
It has been observed that the inetd daemon got killed automatically on both the servers (within a gap of 1 hour) and the users were unable to telnet.

In another K360 server with HP-UX 10.20 the inetd daemon got killed yesterday.

I started the inetd daemon and then they became ok.

I would like to know what could be the possible causes for the inetd daemon being killed automatically.

There is no information of the inetd daemon in the syslog.log. What needs to be checked?

Thanks,
Raje.

[Chris Wilshaw]

It depends on whether the process is being killed, or dying.

If it's being killed, check cron jobs, adn find out if any users with root access have been killing things.

If it's dying, check for the latest inetd patches.

[Sanjiv Sharma_1]

Hi Chris,

Nobody except me has the root access to this servers and hence I don't think someone is stopping the inetd service.

Can you help me to look for the latest inetd patches as I didn't find in the ITRC site.

Thanks,
Raje.

[Chris Wilshaw · Accepted solution]

This link should take you to it.

http://support4.itrc.hp.com/service/patch/patchDetail.do?patchid=PHNE_20747&context=hpux:800:10:20

[U.SivaKumar_2]

Hi,
are you running samba in that server ?

regards,
U.SivaKumar

[Sanjiv Sharma_1]

Hi Chris,
Thanks. I think this patch should help.

Hi Siva,
Yes. Samba is running in all the three servers.
Has Samba got anything to do with this?

Thanks,
Raje.

[S.K. Chan]

There is only one incident that I can relate to you with regards to a frequent dying of inetd. In my case it left behind a core file in root level (/). If I am not mistaken it was caused by a library patch (I can't remember). After I removed that patch it was ok. You might want to check if there is any core file and one more thing.. is your inetd started with the "-l" option ? If not start it with "-l" so that you get inetd logs in syslog.log file, that way it'll help determine the possible causes.

[Sanjiv Sharma_1]

Hi Chan,

I don't find any core file available.

Thanks,
Raje.

[U.SivaKumar_2]

Hi,
It is the problem caused by samba . I have faced it personally. Try stopping the samba in
the problematic server and observe it for some days. I bet that samba is the culprit.

regards,
U.SivaKumar

[Sanjiv Sharma_1]

Hi Siva,

I agree with you but Samba is the most important application running on these servers.
I cannot stop this.

Even if it is a samba problem then what is the solution?

Thanks,
Raje.

[U.SivaKumar_2]

Hi,
Problem is noted with swat enabled samba. SWAT
has this bug.

regards,
U.SivaKumar

[Sanjiv Sharma_1]

Hi Siva,

Thanks. But what is the fix?

Raje.

[U.SivaKumar_2]

Hi,
Only way is to disable SWAT in your samba .

regards,
U.SivaKumar

[Sanjiv Sharma_1]

Hi Siva,

Thanks. But can you tell how can I disable swat in samba.

If you can give me the procedure.

Thanks,
Raje.

[U.SivaKumar_2]

Hi,
Take a copy of /etc/inetd.conf first as backup.
edit /etc/inetd.conf file
you will find a line starting with "swat" .
Remove or comment this line with #.
Give this command.
#inetd -c
This is enough to disable SWAT.

regards,
U.SivaKumar