Link to home
Start Free TrialLog in
Avatar of mandybird44
mandybird44

asked on

network errors.

I installed Solaris2.7 on four machines.
two of them are in the inner network of the firewall and the other two are in the dmz.
I look at /var/log/syslog on one of the machines on the dmz and I found there that message:"Unable to qualify my own domain name...". (by sendmail)
I don't care about the message but my problem is that in this log there were these messages for all other three machines. The IP that appeared there was the real IP followed by two more numbers separated by dots. Can someone please explain this?
Avatar of jlevie
jlevie

Solaris 2.7 uses sendmail 8.9, which in turn is very militant about needing a Fully Qualified Domain Name (FQDN). When it starts it takes the IP of the machine it's running on and tries to look up the name bound to that IP.

If the machines names can be resolved via DNS, and dns services have been properly configured (/etc/resolv.conf and the right "hosts" line in /etc/nsswitch.conf, and a route to the nameserver(s) if they aren't local), you won't get the error. If using dns isn't an option, you can change the host definition in /etc/hosts to have the FQDN as well as the short name, something like:

192.168.0.1      chaos.mystery.com chaos

I'm not sure what your question was about the log file is. Are you asking why the messages from the other machines are there? Or what the numbers mean?
I had the same problem when I first installed Solaris 7 on a Sun box.

A quick call to Sun confirmed the above answer from jlevie above.  If you use the full FQDN the problem and error message in the log file goes away.

The Sunsolve Doc 18391 gives full details on how to cure this problem and an in-depth explanation of it's cause.
ASKER CERTIFIED SOLUTION
Avatar of IainWallace
IainWallace

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
For IainWallace...

Hmm, I could have marked my comment as the answer, but the question has two parts, the second part of which requires input from the questioner to complete the answer.

It's  also not quite proper to effectively use someone else's comment as use your answer
Avatar of mandybird44

ASKER

well thank you but I already know that. What I'm actually asking is the 2 questions jlevie asked in the end of his comments.
Okay, how about pasting one of the error messages into a comment so I can see it. I log slightly different messages when this situation occurs, which isn't surpising as my naming services are undoubtedly set up differently. I can remember seeing that kind of syslog message, but without seeing it again...

The messages from all of the systems are being logged to that system because it's the "loghost" as far as the other machines are concerned. How you get to this configuration depends on how the machines got set up and what kind of naming services are in use (dns, NIS, NIS+). If you can tell me a bit more about the environment I can tell you where to look and how to change it if you want to.
I guess it was because we put as loghost the same computer in each machines /etc/hosts.
10q.
sorry jlevie I wanted to give the points to your last comment but I accidently pushed the button already.
Ah, well... at least it's nice to know that I had the answer.
10x again