Sendmail relaying issue related to EMS and ISEE

[Douglas James Cameron]

Hello,

I have an HP rp4440 running 11.23, and sendmail version 8.11.1 with patch PHNE_31734.

We have our sendmail disabled, to the point where it can only send mail out of the server.
In our aliases file, we have the Exchange e-mail addresses with nicknames for the aliases.

I know we have relaying and mail forwarding, but not through use of .forward files.

We had a power supply (redundant spare) go out. ISEE generated logs, and the only way I knew (as opposed to all of our unix admins normally), was that I was on the server and saw a new mail message come across the command line, going into mailx, I saw a weeks worth of EMS messages. I read the EMS messages that all pertained to the power supply. I immediately opened a case with HP support and told them of the issue. Unfortunately, they can only take so many characters for the description.

I had the power supply replaced the next morning, saw one more EMS message in mailx, but nothing else. It seems HP never even saw the ISEE logs.

Nothing has changed in the cofiguration of sendmail, isee, ems or stm that I can see. The rest of the cluster have the same model, same version OS, smae sendmail dist ver and patch, and seem to have no problem; nothing has changed on our exchange server or firewall, so the questions are:
a. Why did isee not mail anything out
b. Why is sendmail not doing root mail forwarding
c. What is the easiest plan of attack to find out what is going on with the setup?

Thanks in advance,
Doug

[Matti_Kurkela]

How is your Sendmail disabling performed?

If you have changed the SENDMAIL_SERVER setting in /etc/rc.config.d/mailservs to 0, have you set up an alternate means to periodically process the outgoing mail queue? The usual way to do this is to set up a cron job which runs about hourly and executes "sendmail -q".

If you do not do this, outgoing email works well when your network connectivity is OK and the receiving server(s) respond without delay, because sendmail makes one attempt to send the mail when invoked by the sending application. But if something goes wrong on this first attempt, the message is dropped into /var/spool/mqueue.

Normally, the sendmail daemon processes this queue in 30-minute intervals... but if you've disabled the sendmail daemon, this does not happen. If nothing else is done to replace the missing functionality (e.g. the cronjob mentioned above), all the emails that won't succeed at the first try will be stuck forever in /var/spool/mqueue.

Run "mailq" to see if this is the case: if there are old outgoing messages in the queue, this might be what has happened.

You might also look into /var/adm/syslog/mail.log to find out what happens to the ISEE log emails in general.

[Armin Kunaschik]

I don't think you have a sendmail issue here:

a) ISEE does not mail anything from the local host. It uses the HP Website (e.g. europe.europe.hp.com) to generate mail to the configured addresses.
b) root is normally excluded from any redirection. You have to enable this in your aliases file or in your sendmail.cf
c) Are you sure that your host is correctly setup AND _entitled_ for ISEE use? You need to see the entitlement in the web interface http://:5060/start.html
Check logs in /opt/hpservices/log, mad.log is the most important!

My 2 cents,
Armin

[Douglas James Cameron]

HP support had to be involved. It was mostly ISEE needing to be upgraded. Especially after seeing mad.log entries regarding source file errors.

Thank you both for your help, they gave a great starting point.