OpenEMM Support

This case is handled special. As bavd.py is unable to find the DSN for the bounce, but has a rule to mark this as a hardbounce, this is considered as an internal recognized hard bounce and is not written to bounce_tbl. You can edit update.py to write an entry to bounce table by changing line 342: el...

To your problem: Apparently, it doesn't respond to incoming SMTP requests so all messages are dropped. The process to handle incoming mail is semu.py which is the process using port 25. What do you mean exactly by "dropped"? Is the message accepted, but not processed? Or can't the message ...

If the log file is completely empty and the hard disk is not full, then I have no more ideas to fix the issue without having access to the system, sorry.

-- ud

And there is no further information in the sendmail log file?

-- ud

This looks strange. Can you check, if your system is configured to use sendmail at all? For that, please login as user "openemm" and execute these commands: source $HOME/bin/scripts/config.sh $HOME/bin/scripts/smenable.py status This should output either "1" (sendmail is enabled)...

In general, the bounce management should work with the built in SMTP server. Unfortunately your information is a bit spare, so could you please do some checks: Is there any WARNING/ERROR/FATAL message in the semu logfile? A problem in any other logfile, e.g. the update log? Try sending mail to the d...

These functions are originally provide by the termcap library. The parts of SLang referencing these functions are not used in OpenEMM, so I'm a bit surprised, that these leads into a link error. You can try to find out which library provides these functions on your operating system and add them to L...

It looks like mailer.mydomain.net is configured in any way as an alias for the hostname, so the mail is not forward to the bounce management.

Is there any reason for your entry in /etc/alias (and, do you mean /etc/aliases)?

-- ud

We try to avoid false bounce informations due to misconfigured mail servers, so we only try to accept hardbounces which are really proved to be a hard bounce. These are the codes 511 (user unkown) and 512 (domain unknown). To avoid marking your users due to a false bounce message, but still catch th...

If you modify bav.c and insert these two lines just before the inclusion of bav.h: #undef false #udnef true then it should compile. -- ud Same as a unified context diff: --- bav.c.orig 2014-06-16 11:58:04.897116499 +0200 +++ bav.c 2014-06-16 11:58:21.215021164 +0200 @@ -29,6 +29,8 @@ # include <rege...

Looks like some other include file already defines "false", 'true" or "bool_t". So please try these steps: - Change to the source directory for bav.c - Start the command: gcc -E -I../lib -I/home/oe/include -pthread -c -o bav.i bav.c - Look into the file "bav.i" for...

sendmail isn't writing directly to /var/log/maillog, it uses the syslog mechanism. So this looks like something is wrong with your syslog configuration.

-- ud

There is only a limited support for graylisting detection; in this case the bounce is ignored and not explicit recorded. All others are recorded as softbounce (if matching the remaining rules for softbounces), so all bounces from graylisting, which are not identified by the available logic is counte...

OpenEMM tries to filter all automatically generated mails so only human written mails should be forwarded by the bounce management. In your example, the mail is recognized as an automatic mail (due to sender <>). These mails are swallowed by the bounce management.

-- ud

The logfiles are found in /home/openemm/var/log and herein you should have a look into the file <timestamp>-<hostname>-semu.log for further information.

-- ud