manually executed script

[Jonathan Grymes]

After I execute the script (as root) it returns Mail sent.
# ./CheckFileSystems

Sending mail...
Mail sent!
#

When I check /var/mail/root I see the following entry:
for root; Thu, 29 Nov 2007 07:21:00 -0500 (EST)
Date: Thu, 29 Nov 2007 07:21:00 -0500 (EST)
From: root@fgexsh41.vad.ws-gex.com
Message-Id: <200711291221.lATCL0612170@fgexsh41.vad.ws-gex.com>

Cron: Your job did not contain a valid audit ID.
See your system administrator.

I was able to send mail via ELM so I know that is working.

Any thoughts?

Thanks,

[Steven E. Protter · Accepted solution]

Shalom,

This happens sometimes when converting systems with cron authorization to trusted mode.

Fix is found:
http://docs.hp.com/en/J4269-90041/J4269-90041.pdf

Its kinda buried in that document.

Here is how cron does it:
http://www.auditnet.org/docs/HPUX.doc
http://www.adager.com/VeSoft/HpUxSecurityConcerns.html

SEP

[Dave Hutton]

I'm not totally following this. Are you saying when you run the script manually everything is fine?

roots email you show indicates to me that it was something from cron. Anything in /var/adm/cron/log around 7:21?
Usually if cron has an issue it will email root or the user it runs as.

Anything when you type crontab -l?

The only few hits I got when searching for that error is about tsconverting your system to trusted systems. Was that recently done?

What does cat /tcb/files/auth/r/root look like?

[Jonathan Grymes]

The script is supposed to generate an email to send the results. I never receive it. The box was recently put in audit mode. I know ELM is working.

var/adm/cron/log:

> CMD: sleep 15;/etc/opt/resmon/lbin/mon_EMSHAProvider_state.sh
> root 11939 c Thu Nov 29 07:19:00 EST 2007
< root 11939 c Thu Nov 29 07:19:00 EST 2007 rc=1
> CMD: /opt/sfm/bin/restart_sfm.sh
> root 12112 c Thu Nov 29 07:20:00 EST 2007