Greetings Mike:
We've seen this happen when spamd dies or otherwise gets overloaded.
You may want to look at the "max children" option; you may have no value (and therefore want to try it; we use 10 on busy mail servers and that appears to work ok) or a high value (for which you may want to lower).
You may also want to check that your /tmp partition is not full, and that the socket file used by SA can be created (upon service start) and deleted (upon service stop) without problems (i.e. directory and file ownership, group ownership, and permissions).
Thank you.
At 12:21 PM 3/21/2005, you wrote:
I am running a global SpamAssassin 3.0.2 (spamd and spamc) via procmail on a Solaris box (spamd and spamc running on the same server). During periods of high load, I get the following messages in the syslog.
spamc[416]: [ID 702911 mail.error] connect(AF_INET) to spamd at 127.0.0.1 failed, retrying (#1 of 3): Connection timed out
I think these are what is causing some messages to come through unscanned. Is there a way I can increase the number of retries and the timeout setting for spamc? I know I really just need to throw a dedicated server at it, but I would like to know how to keep this from happening.
-- Mike Robinson <[EMAIL PROTECTED]>
