chance@sfsuvax1.sfsu.edu (Jon Naumann) (06/27/91)
I am a novice Sys Admin with an old HP9000s300 running HP-UX 7.0 that has just been installed from scratch. Users can send mail to each other just fine, but when other hosts attempt to send mail they get errors like : ----- Transcript of session follows ----- While connected to foghorn.sfsu.edu [130.212.10.100] (tcp): >-> HELO foghorn <<< 553 foghorn I refuse to talk to myself 554 <chance@foghorn.SFSU.EDU>... Service unavailable ----- Unsent message follows ----- From: chance@sutro.sfsu.edu (Jon E. Naumann) To: chance@foghorn.SFSU.EDU chance@foghorn... Connecting to foghorn via ether... mail exchanger is foghorn.sfsu.edu Trying 130.212.10.100... connected. 220 foghorn HP Sendmail (15.11/15.6) ready at Thu, 27 Jun 91 07:17:48 pdt >>> HELO sutro.SFSU.EDU 250 foghorn Hello sutro.SFSU.EDU, pleased to meet you >>> MAIL From:<chance@sutro> 250 <chance@sutro>... Sender ok >>> RCPT To:<chance@foghorn.SFSU.EDU> 250 <chance@foghorn.SFSU.EDU>... Recipient ok >>> DATA 354 Enter mail, end with "." on a line by itself >>> . 250 Ok >>> QUIT 221 foghorn closing connection chance@foghorn... Sent I'm perplexed as the verbose mail from the remote host says that it sent the mail, but then the HP sends it back with the criptic "i refuse to talk to myself" message. Is the system just schizo, haunted by Rod Serling, or suffering from an inadequate configuration ? Does anyone have any ideas ? I'd certainly appreciate any comments ... e-mail directly to chance@sfsu.edu or post. Thanks in advance ...
cricket@hpcc01.HP.COM (Cricket Liu) (06/28/91)
I am a novice Sys Admin with an old HP9000s300 running HP-UX 7.0 that
has just been installed from scratch. Users can send mail to each
other just fine, but when other hosts attempt to send mail they get
errors like :
----- Transcript of session follows -----
While connected to foghorn.sfsu.edu [130.212.10.100] (tcp):
>-> HELO foghorn
<<< 553 foghorn I refuse to talk to myself
554 <chance@foghorn.SFSU.EDU>... Service unavailable
Lovely message, no? What it's trying to tell you is that foghorn received
the mail, didn't recognize it as being addressed to a local recipient, and
tried to forward it on. When it popped open its own SMTP port and realized
its mistake (by noticing the hostname in the HELO), it shut down the
connection to avoid a mail loop.
I'm guessing that sendmail isn't canonicalizing "foghorn" to
"foghorn.sfsu.edu" when it freezes the sendmail.cf. Is the host configured
to use name service? If it weren't, that would do it. Check that
/etc/resolv.conf reads something like:
domain sfsu.edu
nameserver 130.212.10.102
nameserver 130.212.15.230
Heck, it may be that the machine just booted at a time when no name servers
were reachable, so the canonicalization step failed. In that case (name
service is configured and working), just kill the sendmail daemon
(/usr/lib/sendmail -bk), refreeze (/usr/lib/sendmail -bz), and restart it
(/usr/lib/sendmail -bd -q30m, or some other queue interval).
If that's not it, drop me a line. I'd be happy to help you out if I can.
cricket
hostmaster@hp.com