Re: "Configuring header/body checks... from outside users only" && main.cf

2014-02-14 Thread Wietse Venema
d...@powerstandards.com: > Hi, > > (I believe I have found the problem here and it is unrelated to postfix, > but I would like to confirm a couple of things that are a bit > hazy in the docs. After googling a lot yesterday, this may be a > useful find for someone in the future.) > > I got header

Re: "spontaneous" re-appearance of .db (and source text!) files

2014-02-14 Thread Wietse Venema
d...@powerstandards.com: > Hi Wietse, > > This is a syslog snapshot of the situation a few seconds earlier: > * postmap processes dying because of a missing file (remember, there is no > mention of the db file in the current configuration) > * other postmap processes creating/populating the db fil

"Configuring header/body checks... from outside users only" && main.cf

2014-02-14 Thread dave
Hi, (I believe I have found the problem here and it is unrelated to postfix, but I would like to confirm a couple of things that are a bit hazy in the docs. After googling a lot yesterday, this may be a useful find for someone in the future.) I got header checks working, but ran into the Catch-22

Re: "spontaneous" re-appearance of .db (and source text!) files

2014-02-14 Thread dave
Hi Wietse, This is a syslog snapshot of the situation a few seconds earlier: * postmap processes dying because of a missing file (remember, there is no mention of the db file in the current configuration) * other postmap processes creating/populating the db file "N" times, ergo the duplicate ent

Re: "spontaneous" re-appearance of .db (and source text!) files

2014-02-14 Thread dave
Hi Wietse, Let me start by saying thank you for this software!! > Why are you running postmap every 5 seconds? I am not (as far as I know--well, clearly I am but I don't know why), at least not on purpose. I did not notice it myself until I started trying to debug this issue. The output you re

Re: "spontaneous" re-appearance of .db (and source text!) files

2014-02-14 Thread Wietse Venema
d...@powerstandards.com: > Feb 14 08:47:11 P1234567 mail.warn postfix/postmap[16487]: warning: > /etc/postfix/sender_access2.db: duplicate entry: "xx...@yyy.com" > Feb 14 08:47:13 P1234567 mail.warn postfix/postmap[16499]: warning: > /etc/postfix/sender_access2.db: duplicate entry: "xx...@yyy.com"

"spontaneous" re-appearance of .db (and source text!) files

2014-02-14 Thread dave
I am seeing a very strange effect. I documented it from the command line below. I am running on an embedded system. What we are trying to do is whitelist a very short list of recipients. I am trying to set up some header and address verification checks, and will soon have written about that in a

cyrus SASL problems with DIGEST-MD5

2014-02-14 Thread Matthias Leopold
hi, i'm having trouble with switching our postfix servers from dovecot SASL to cyrus SASL for SMTP-AUTH. i'm using "auxprop_plugin: sql" with "sql_engine: pgsql". the username for authentication is the full email address: u...@domain.tld. the "sql_select" query uses the '%r' macro like this:

Re: Domain lookup fails for exactly one domain

2014-02-14 Thread Viktor Dukhovni
On Fri, Feb 14, 2014 at 09:39:00AM +0100, Sven Schwyn wrote: > > As "root" or as "postfix"? > > I can resolve the domain as both. > > > Is this domain using a dedicated master.cf transport? Is that > > transport's smtp(8) delivery agent chrooted? > > The master.cf is untouched as it comes with

Re: Domain lookup fails for exactly one domain

2014-02-14 Thread Sven Schwyn
Victor wrote: > As "root" or as "postfix"? I can resolve the domain as both. > Is this domain using a dedicated master.cf transport? Is that > transport's smtp(8) delivery agent chrooted? The master.cf is untouched as it comes with the Gentoo package. Nothing is chrooted. (It's a send-only mac

"Satellite" config vs null client config

2014-02-14 Thread Daniel Pocock
The Debian package of Postfix offers a "Satellite system" configuration (a list of all the Debian configs at the bottom of this email) I'm not sure if this is meant to emulate the null client configuration or be something else Anyway, I opened an issue for it in the Debian bug tracker http: