Noah writes:
any clues why mailman keeps dumping the following error trace in /usr/local/mailman/logs/error ? Happens about every hour or so or sometimes minutes apart.
Probably for the same reason you repeatedly post to Mailman Users: it wants attention to its pain.
For future reference, the folks (not me! I show up when I feel like it) who monitor this list are knowledgeable and conscientious, in the extreme, and it is *very* rare that posts get dropped on the floor. Typical response times are in the range 30 minutes to 36 hours, depending on random factors.
Check that your post has appeared in the archives, and if so, give it a day at least. Don't waste the interval; read the FAQ[1], search the list archives (a search for the error message, in this case "Ignoring unparseable message", is strongly recommended), collect more log data, etc. If your post doesn't appear within a day, you may wish to send mail to the list administrator. It's possible that your post got inadvertantly snagged by a spam filter, and you'll need help getting past that.
Dec 26 08:27:28 2006 (75889) Ignoring unparseable message: 1166121349.6141329+c811d30798575d4ca57b9a0552a1d394a95c8ff1
The hash code above identified the problem message. I believe you will find a file by that name (with extension .pck) in the "qfiles" directory of your Mailman data area, and if you move it somewhere else, Mailman will stop complaining. It happens regularly because Mailman tries to process the files in that directory on a regular basis, until all deliveries scheduled for it are complete.
It's possible that you are the target of a spambot, in which case you may not find the file (because it's been shunted to the "shunt" directory), and you will see that all the message IDs are different in the logs, because the spambot is sending a new copy at hourly intervals.
Dec 26 08:27:28 2006 (75889) Uncaught runner exception: No terminating boundary and no trailing empty line
You've got a badly-formed MIME message in the queue. It could be spam, it could simply be the product of a Microsoft mail client, it could be broken by a relay MTA, or it could be that you ran out of diskspace and it got truncated (seems unlikely, given the care that Postfix and Mailman take to not do such things).
Use the bin/show_qfiles script to examine the contents of that file. Simply changing to the mailman data area (on my Debian system it's /var/lib/mailman, YMMV) running bin/show_qfiles without arguments will give you a usage message.
Footnotes: [1] The whole thing, both the static one and the FAQ Wizard. If it's not important enough to read the whole FAQ, which isn't as much fun as a Sidney Sheldon novel but is also much shorter, it probably can wait a day or a week.