tony@mpg.phys.hawaii.edu (Antonio Querubin) (05/28/91)
The hostname for our system just changed and I would like it to accept mail for both the old and new hostnames during a transition period. I thought this was simply a matter of putting the following rule into sendmail.cf: R$*oldname$* $1newname$2 I put this into several rulesets and it didn't really work except for local mail addressed to the old hostname. When a remote host makes a connection to our system I get an error message generated to postmaster: <<< 553 mpg.phys.hawaii.edu host name configuration error 554 <tony@uhcmtg.phys.hawaii.edu>... Service unavailable So I know the remote machine is making the SMTP connection but the local system is rejecting the mail. I'm running SUN's sendmail with SunOS 4.0.3. Anyone got any clues for me on making this work? Perhaps I'm just putting the rule into the wrong ruleset? Please respond here rather than by e-mail as it will be a while before our local nameservers are updated with the new hostname. Thanks! -- Antonio Querubin tony@mpg.phys.hawaii.edu / ah6bw@uhm.ampr.org / querubin@uhunix.bitnet
rickert@mp.cs.niu.edu (Neil Rickert) (05/29/91)
[Followup in case the email message was not received] In article <13228@uhccux.uhcc.Hawaii.Edu> tony@mpg.phys.hawaii.edu (Antonio Querubin) writes: >The hostname for our system just changed and I would like it to accept mail >for both the old and new hostnames during a transition period. I thought >this was simply a matter of putting the following rule into sendmail.cf: > >R$*oldname$* $1newname$2 > >I put this into several rulesets and it didn't really work except for local >mail addressed to the old hostname. When a remote host makes a connection to >our system I get an error message generated to postmaster: When the address works correctly on locally originating mail, but fails on mail from elsewhere, always look to see if the sendmail daemon has been killed and restarted since modifying 'sendmail.cf'. Otherwise the daemon may still be using the old configuration. There is no way to persuade the daemon to re-read its configuration dynamically. You have to kill and restart. Look in /etc/rc* to find the exact startup command. -- =*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*= Neil W. Rickert, Computer Science <rickert@cs.niu.edu> Northern Illinois Univ. DeKalb, IL 60115 +1-815-753-6940
tony@mpg.phys.hawaii.edu (Antonio Querubin) (05/29/91)
In article <1991May29.000921.9277@mp.cs.niu.edu> rickert@mp.cs.niu.edu (Neil Rickert) writes: > > When the address works correctly on locally originating mail, but fails on >mail from elsewhere, always look to see if the sendmail daemon has been killed >and restarted since modifying 'sendmail.cf'. Otherwise the daemon may still >be using the old configuration. There is no way to persuade the daemon to >re-read its configuration dynamically. You have to kill and restart. ..... Yes that pretty sums up what I had forgotten to do as several people pointed out to me. When I got the first reply to my question I wanted to beat my head against the wall... Thank you very much to the folks who responded! -- Antonio Querubin tony@mpg.phys.hawaii.edu / ah6bw@uhm.ampr.org / querubin@uhunix.bitnet
Craig_Everhart@TRANSARC.COM (05/29/91)
(I'm sending this to comp.mail.sendmail because it's unclear how to get mail to the sender, given the complaint!) Which is the old name and which is the new? At the moment, mpg has an MX record pointing to uhcmtg, and uhcmtg is listed as a CNAME for mpg (all correctly translated to fully-qualified domains). You can't have both of these. The target for an MX record needs to be a name that isn't a CNAME. Craig