(Please cc me on any replies - I am not currently subscribed to the list) I've been running spamassassin/spamd at my site quite happily for quite a while. Recently, the load on my one incoming smtp server has gotten quite high during peak delivery times from spamd processes.
I already have a handfull of completely identical "satellites" that nfs mount home directories from the central smtp server and serve out imap, pop, ssh, etc. through the use of a linux virtual server. To lighten the load on the main smtp server, I've been contemplating pointing spamc over the network to each of the satellites, and eventually to the head of the LVS. The satellites are all redhat-7.2 based machines, with current updates, and some local customization. I've installed spamd (2.60) through CPAN (install ExtUtils::MakeMaker File::Spec Pod::Usage HTML::Parser DB_File Net::DNS Mail::Audit Mail::Internet Net::SMTP Digest::SHA1 Net::Ident Mail::SpamAssassin) onto all of them, and started spamd with these options "-i 0.0.0.0 -A <spamc_server_ip> -d". I've been using perl, compiled from source (sh Configure -de && make && make test && make install), versions 5.6.1, 5.8.0, and 5.8.1. I don't use razor or pyzor. spamd runs fine on the machines for a while, but then dies & outputs this error to syslog: spamd[1637]: Use of uninitialized value in subroutine entry at /usr/local/lib/perl5/5.8.1/i686-linux/Socket.pm line 370. It takes differing amounts of time to die each time. I have one identical satellite that isn't in production and runs spamd just fine - without dying at all. In trying to figure out what was different with this machine than the others, I finally came to the realisation that I don't monitor the machine. Further troubleshooting led me to begin to think that the periodic nmap scan I run against the production satellites might be causing spamd on them to die. In testing, I've found that the production spamd satellites do in fact die once in a while from a plain nmap scan with no arguements, but often work just fine. I've since disabled the nmap scans, and spamd hasn't crashed on any of the satellites. Has anyone else seen similar problems? or care to confirm my suspicions that the nmap scans could be causing spamd to die? Thanks in advance, -Mark -- Mark T. Valites Unix Systems Analyst CIT - SUNY Geneseo >--))> >--))> ------------------------------------------------------- This SF.net email is sponsored by: The SF.net Donation Program. Do you like what SourceForge.net is doing for the Open Source Community? Make a contribution, and help us add new features and functionality. Click here: http://sourceforge.net/donate/ _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
