On Wed, Jun 02, 2010 at 05:14:45PM +0200, Proniewski Patrick wrote: > So it appears that the connection between MAILGW and LB is not always > properly closed. Am I wrong?
http://www.postfix.org/postconf.5.html#smtp_connection_cache_on_demand http://www.postfix.org/CONNECTION_CACHE_README.html http://www.postfix.org/postconf.5.html#connection_cache_ttl_limit > smtp_destination_concurrency_limit = 10 A bit low, especially when most traffic goes to the same place. I would recommend the default of 20 or even 50 in some similar cases. Increasing the concurrency limit will likely reduce the backlog and demand for cached connections. > smtp_destination_recipient_limit = 15 This too is likely counter-productive. Decreasing envelope splitting will reduce the number of messages and thus demand for cached connections. > tcpdump output: > 15:51:22.010668 > 220 co7.domain.tld SMTP ready > 15:51:22.010746 > EHLO mailgw.domain.tld > 15:51:22.011256 > 250-co7.domain.tld > 250-AUTH CRAM-MD5 DIGEST-MD5 GSSAPI LOGIN PLAIN > 250-STARTTLS > 250 8BITMIME > 15:51:22.011337 > MAIL FROM:<sen...@mail.domain.tld> BODY=7BIT > 15:51:22.011736 > 250 OK > 15:51:22.011803 > RCPT TO:<xxxxx...@mail.domain.tld> > 15:51:22.021497 > 250 OK > 15:51:22.021574 > DATA > 15:51:22.022099 > 354 Start mail input; end with <CRLF>.<CRLF> > 15:51:22.022211 . 120:1568(1448) > 15:51:22.022224 . 1568:3016(1448) > 15:51:22.022235 P 3016:4216(1200) > 15:51:22.022786 P 5664:5721(57) > ... > . > 15:51:22.062987 . ack 5721 > 15:51:22.275092 P 197:205(8) > 250 OK > 15:51:22.374644 . ack 205 Postfix caches the connection and does not send "QUIT". > 15:51:24.276369 F 5721:5721(0) Postfix closes the connection two seconds later. > 15:51:24.276858 P 205:265(60) > 221 co7.domain.tld service closing transmission channel The LB or the server behind it violates SMTP by sending an out-of-turn SMTP reply. > 15:51:24.276910 R > 15:51:24.276916 F 265:265(0) > 15:51:24.276930 R 4094209693:4094209693(0) win 0 Postfix is already gone. This is harmless, but the either LB or custom mail-server behind it mishandles EOF by sending an out-of-turn reply. If you don't want connection caching, turn it off. -- Viktor.
