Milter to error state observation

[Bob Falkenberg]

Hello Folks, 


Background

I have been running the milter on a server that has been doing an average of
25,000 messages a day for 6 months now.  We are using a Redhat 8 server
configured pretty much verbatim off the www.peregrinehw.com website with
Sendmail 8.12.9, SA 2.44 and the 1.30a milter.  I have a script that watches
the mail log for "milter to error state" messages restarts the milter and
spamassassin if it gets to many errors.

Observation
New software Dev Group here is writing can easily swamp DNS server with
massive numbers of queries.  As the DNS server gets bogged down it takes
longer and longer to respond and finally causes server to stop responding...
DNS admin had been trying to balance the load and started using the same
server that SA was using to do internal DNS lookups.  DNS server bogs down
SA starts throwing milter errors.  Admin restarts DNS sever and milter
errors stop..... Server bogs down again... Milter errors return..... 

Might be something to investigate.... I am am using a different DNS server
that's not involved in that project and not a peep out of the check milter
script.

Bob F.