Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!cs.utexas.edu!samsung!munnari.oz.au!uhccux!catfish!jwright From: jwright@cfht.hawaii.edu (Jim Wright) Newsgroups: comp.sys.hp Subject: HP sendmail doesn't work (with free bugfix!) Message-ID: Date: 2 Sep 90 07:24:53 GMT References: <59169@bbn.BBN.COM> <1340137@hpclscu.HP.COM> Sender: news@uhccux.uhcc.Hawaii.Edu Lines: 45 shankar@hpclscu.HP.COM (Shankar Unni) writes: >Please take the time to call HP support and let them know what problem you >are having, so that they can take a crack at isolating the problem. What a perfect opening, I can't pass it up. We've done all the right things. Sendmail stopped working for us with the 7.0 upgrade. We put in a call to the response center, and sometime during the next few months it was quietly closed. So we opened a new call, and sent email to various contacts in HP. Still no solution. PROBLEM 1: Sendmail simply does not work at all with /usr/lib/aliases. Long mournful tail of my woes omitted. After about a month on the second call, we got an answer "back from the factory". Seems that the sendmail.cf that was shipped was broken. OK, I can accept that. What toasts my tortilla is that the "factory" knew all about this, while I tried to struggle through the response center. What's the point of a service contract? So, at the risk of giving away thousands of dollars in free support... At about line 780 in the "HP supported sendmail.cf" you will find an "On" setting. This should be modified to read "Onfalse". You will then need to reboot, refreeze the configuration files, and run newaliases. # check right sides of aliases on alias db initialization Onfalse PROBLEM 2: The fixed sendmail will not resolve through two aliases. This one has gone "back to the factory". Oh shit, not again. So my plea to the net: why does this not work? An example below. wright: jwright jwright: jwright@quonset jim: jwright "quonset" is a cluster server. If I go to another cluster and send mail to either wright or jim, sendmail resolves the first alias and sends mail to /usr/mail/jwright _on_that_cluster_!! Wrong!! I can use the command "/usr/lib/sendmail -v -s test jim < /dev/null" and watch the progress. On any of our Suns, this works like a charm and mail goes where it should. On the HPs... HELP!! -- Jim Wright jwright@quonset.cfht.hawaii.edu Canada-France-Hawaii Telescope Corp.