Folks -- are you using the -m switch to limit how many spamds can run? If you don't use that, and you're using an MTA that also does not limit concurrent local delivery commands, like sendmail, this will be an issue. Try the -m switch.
On some setups (notably Sendmail/Procmail) the -m option of spamd can be even worse than the original problem.
It ends up causing a huge backlog of sendmail and procmail processes as the spamc clients *wait* for a free spamd process. This alone can consume enough memory to bring the machine down.
Also, if the spamc processes time out waiting for spamd to answer, they return an error code to procmail which depending on your script can trigger the lost "F" bug of procmail or other problems :/
It's better to try and limit the concurancy at the MTA or MDA level. My hack for the problem is to return temporary failure in my local delivery script before procmail is even called, if there are too many delivery processes going on.
Another way to do it would be to check at the beginning of the procmail script, count how many spamd processes are active (is there a tidier way to find this out from spamd directly rather than using ps ?) and if its over a set limit, exit with EX_TEMPFAIL. A workaround like that could be used by anyone calling spamc from procmail...
I might have a tutu around later and see if I can work out a procmail recipie to do that...
Regards, Simon
------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
