Re: [Mailman-Users] Uhhuh, my GID blues goes on... :( - Case SOLVED!
On Mon, 2005-11-28 at 10:23 +0200, Niemi Hannu wrote: > > This would seem to say that postfix is using two different > > GIDs to invoke the wrapper depending on the list (or, if this > > is the case, on where the list's aliases come from). I think > > you need to carefully look over your Postfix configuration > > and see if it might be getting mailman aliases from two > > different places and using different GIDs for the different places. > > Aargh.. I feel dumb, dumber, dumbest... > > Though, I still don't 100% grab it, I managed to solve the problem, > which was MUUUCH silleer I did even anticipate. I had read the postfix > configuration many times through but forgot to check the postfix > aliases-file, which had all the vintage lists in it. After removing the > lists form the aliases file everything started to act logically. > > I did put the lists in there when setting the listserver fast up after > the previous server crashed (physiaclly). Afterwards I added the > mailman-generated aliases-file into postfix aliases and just forgot to > delete the mailman-related aliases from the aliases file. As the lists > worked all right (through the probably primary aliases file) I didn't > notice any problems during that time. The problem just manifested itself > as now, as we needed a new list. > > What is strange is that the wrappers in the aliases file was just the > same as in the mailman aliases file, so it still beats me, why it didn't > work. But, let's hope this cured the problem until forever! :) The reason is because of a Postfix feature. Postfix will by default, although it is configurable, run external commands found in an alias file under the uid/gid of alias file it found the alias in. Thus you can have postfix run external commands using a specific uid/gid by partitioning your aliases into separate files, all of which share a common uid/gid requirement. If you keep all the mailman aliases in an mailman only alias file with mailman specific uid/gid then when those aliases instruct postfix to run a command (e.g. the wrapper) then it will do so with the uid/gid of the mailman alias file. In your case you ended up with a mixed bag of aliases and depending on which bag the alias was found in first you got the uid/gid of that bag, hence the seemingly inconsistent behavior, which was in fact very consistent. -- John Dennis <[EMAIL PROTECTED]> -- Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/ Unsubscribe: http://mail.python.org/mailman/options/mailman-users/archive%40jab.org Security Policy: http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.027.htp
Re: [Mailman-Users] Uhhuh, my GID blues goes on... :( - Case SOLVED!
Niemi Hannu wrote: > >What is strange is that the wrappers in the aliases file was just the >same as in the mailman aliases file, so it still beats me, why it didn't >work. But, let's hope this cured the problem until forever! :) The problem wasn't with the wrapper per se or with which wrapper was invoked as there was only one. The problem was that Postfix invokes the wrapper as the user/group of the owner of the aliases.db file that the alias came from. In your case, you had two different aliases/aliases.db files with different owners, each containing some mailman aliases, so the wrapper wasn't being invoked with a consistent group. -- Mark Sapiro <[EMAIL PROTECTED]> The highway is for gamblers, San Francisco Bay Area, Californiabetter use your sense - B. Dylan -- Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/ Unsubscribe: http://mail.python.org/mailman/options/mailman-users/archive%40jab.org Security Policy: http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.027.htp
[Mailman-Users] Uhhuh, my GID blues goes on... :( - Case SOLVED!
> This would seem to say that postfix is using two different > GIDs to invoke the wrapper depending on the list (or, if this > is the case, on where the list's aliases come from). I think > you need to carefully look over your Postfix configuration > and see if it might be getting mailman aliases from two > different places and using different GIDs for the different places. Aargh.. I feel dumb, dumber, dumbest... Though, I still don't 100% grab it, I managed to solve the problem, which was MUUUCH silleer I did even anticipate. I had read the postfix configuration many times through but forgot to check the postfix aliases-file, which had all the vintage lists in it. After removing the lists form the aliases file everything started to act logically. I did put the lists in there when setting the listserver fast up after the previous server crashed (physiaclly). Afterwards I added the mailman-generated aliases-file into postfix aliases and just forgot to delete the mailman-related aliases from the aliases file. As the lists worked all right (through the probably primary aliases file) I didn't notice any problems during that time. The problem just manifested itself as now, as we needed a new list. What is strange is that the wrappers in the aliases file was just the same as in the mailman aliases file, so it still beats me, why it didn't work. But, let's hope this cured the problem until forever! :) Thank you all and especially Mark for good suggestions Best regards hannu -- Mailman-Users mailing list Mailman-Users@python.org http://mail.python.org/mailman/listinfo/mailman-users Mailman FAQ: http://www.python.org/cgi-bin/faqw-mm.py Searchable Archives: http://www.mail-archive.com/mailman-users%40python.org/ Unsubscribe: http://mail.python.org/mailman/options/mailman-users/archive%40jab.org Security Policy: http://www.python.org/cgi-bin/faqw-mm.py?req=show&file=faq01.027.htp
