mail problem

[Bill Hassell]

This may be due to newer versions of sendmail that try to validate hosts--sendmail bypasses /etc/nsswitch.conf and follows the policy set in the /etc/mail/service.switch file. Set the two values to:

hosts files dns
aliases files

sendmail is getting fairly convoluted about hostname resolution and mail delivery.

And then there are the MX records (or not) in your DNS server. When sendmail gives up on a hostname, see how it will figure out where to send the mail for a given destination:

nslookup -type=mx infineon.com

[mohan singh]

Hi All ,

Thanx for your kind help , Filnaly problem is resoved.

I copied /etc/mail/*.* to this sever and change the host related entry ..

and stop sendmail and start.

It works.

Thanx once again for your kind and sincere support.

Mohan

[Muthukumar_5]

hai mohan,

IT is great to know that your problem is solved. And good to boost responders with points allocation.

Keep posting questions.

Take care.

[Ahmed_41]

Dear All,

i have the same error mesg, all my nsswitch file is configuered to use files.

my situation is a bit different, i am sending to a local mailserver, and till now i am unable to send to it and i am having the same error mesg here you are talking about.

i have in hosts defined the other mailserver IP with hostname.
i dont know what to do to solve this probelm myself.
can someone help me please

[Marco Rojano]

Moa, can you help me?? I have the same problem that you... "Name server timeout"..etc.

You resolved this problem... can you say me, what exactly you did?? pass by pass??

please.. help me

thanks a lot
Mark

[Steven E. Protter]

Mohan needs to assign some points to those that helped him.

Marco Rojano:

You should start your own thread, reference this one with a link and reward those that help you with points.

Note: Zero points for this post.
Zero points for this post.
Zero points for this post.
Zero points for this post.

I'm serious.

I'm advising you to start your own thread because this one has been sitting around for a while and its stale. No telling when or if the author will get back to you with an answer.

Good ettiquette usually demands that when one solves a problem the post the solution up making the thread more useful when it comes up on an itrc search.

SEP

[Mike Smith_33]

I just finished resolving an issue that is basically identical to what was posted here. I found this post in researching my problem and I thought what I found might help someone.

Problem:
Email to remote domains worked great. Email to my local domain would go into deferred status. I went through all the checks of sendmail.cf, sendmail.cw, resolv.conf, etc. . .One difference for me, two hpux boxes began to exhibit this on the same day. The emails would sometimes flow and then usually a bunch at once.

We set up network sniffer traces and we could see data flowing back and forth between the systems. One thing that stood out was the large number of nameservers being returned by
dns.

After much data review and checking and double checking HP support was able to point that the initial request came as a udp packet. The nameserver list that came back was too large for the udp packet so it was being truncated. Sendmail would then resend the request as tcp and sometimes this would hang for minutes. This got the network and windows guys to checking some things and they found several routes (4) that were invalid. We also found 15 nameservers in the nameserver list that were no longer available. After these lists were cleaned up mail began to flow with no problem.

We believe the issue was that sometimes dns returned a bad nameserver resulting in the mail not flowing. I also think that by reducing the number of nameservers we are not running into the udp truncations issues as well.

Hopefully this quick summary may help someone .