http://bugzilla.spamassassin.org/show_bug.cgi?id=3828





------- Additional Comments From [EMAIL PROTECTED]  2004-10-04 09:46 -------
well, i can only find this happening in 1 location right now, and really the
only difference is this location runs a spamd cluster, and dns round robins the
requests...  spamd is invoked like this:

 /usr/bin/spamd --syslog-socket=none -A 127.0.0.0/8,24.225.0.0/24 -i 0.0.0.0 -
q -x -m 15 --max-conn-per-child 75

and spamc clients connect from remote hosts like this...

/usr/bin/spamc -dspamd -H -p783 -t30 -x -u [EMAIL PROTECTED]

the differences between this setup and all other setups i have that run just
fine are:

1) -dspamd vs -dlocalhost
2) -H flag on spamc call
3) -A flag is 127.0.0.0/8,24.225.0.0/24 vs just 127.0.0.0/8

i thought maybe it was the network infastructure causing it since i have only
seen this happen at a single location, but there doesnt appear to be anything
wrong here.  The 3 MTA's are plugged into the same switch that the spamd cluster
is on, so there is no routing or anything involved here.

i'm still trying to get better debug of it,  its very difficult to do.  i have
written some spamd child timeouts into the spamd daemon that appear to be
working good also.  prevents me from having to restart spamd now to regain mail
flow.   i'll try and clean it up and add a patch for review if anyone is
interested... however, this is not the answer to fixing this problem, but may be
the answer to help prevent runaway children that we have seen in the past.

d





------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

Reply via email to