I haven't posted in a little while. Mimedefang runs so smoothly and I haven't been fast enough to answer questions I know. But now I am puzzled.
We've moved from Solaris to SE Linux RHEL 6, which means we've moved from compiling software to using packages. I am trying to go with the flow. I know it's Sendmail 8.14.4 and Mimedefang 2.78. We had a Spamassassin incompatibility. We have used the PDFInfo.pm plugin for years and configured the new Spamassassin in the .pre file to use it, but the Spamassassin package from Yum does not include PDFINfo.pm (huh?) so not surprisingly we got this: Slave 11 stderr: plugin: failed to parse plugin (from @INC): Can't locate Mail/SpamAssassin/Plugin/PDFInfo.pm in @INC (@INC contains:... We'll deal with that. What I am writing about is Mimedefang's followup to that, which was: sendmail[3600]: u4N5l72x003600: milter_sys_read(mimedefang): cmd read returned 0, expecting 5 sendmail[3600]: u4N5l72x003600: Milter (mimedefang): to error state Thanks to Gmail's practice of sending to one RCPT at a time, we had over a dozen identical messages (same Message-ID). They all got the PDFInfo.pm error but the first ones continued and were sent out after about 4 seconds. After those slow successes, the later ones failed; for those sendmail logged the 2 errors shown, and at that same time, the multiplexor died, and senders got "Please try again later", logged 7 or 8 seconds after the "from=" log line. I see from archived messages (nine years ago!) this could be caused by mismatched milter libraries, sendmail and mimedefang. I would like to believe that the Red Hat people have sendmail and mimedefang linked to the same library, although I am always an optimist. Is there any other cause we can look for? I should add that that we went a few weeks on the new system without ever seeing this. Something unusual happened. My main concern is to avoid having the multiplexor die. I have of course removed PDFInfo from the Spamassassin configuration, pending testing and installing the plugin in the new environment-- so that particular cause will not recur. --Joseph Brennan Columbia University _______________________________________________ NOTE: If there is a disclaimer or other legal boilerplate in the above message, it is NULL AND VOID. You may ignore it. Visit http://www.mimedefang.org and http://www.roaringpenguin.com MIMEDefang mailing list MIMEDefang@lists.roaringpenguin.com http://lists.roaringpenguin.com/mailman/listinfo/mimedefang