dccm causing tempfails

Bart Dumon bartdu@bsp.scarlet.be
Fri Sep 1 14:15:37 UTC 2006


Hi,

Did anyone ever run into a problem with dccm where sendmail would be
tempfailing a lot of incoming connections? The circumstances are a
bit unclear, but it does seems to happen under at least a decent
load on the box. It does not happen when dccm is disabled in 
sendmail.cf.

We have a fair share of incoming mail that gets autoresponded or
relayed further internally (both through procmail) for other purpose 
and at that point we're building large queue's for the MTA if dccm is 
being used. We're currently taking in mail with 6 boxes, 4 of them 
have a large clientmqueue, 2 of them have an empty clientmqueue, those 4 
are running dccm, the other 2 aren't.

Running /xxxxx/spool/clientmqueue/k81Dhg4T011984 (sequence 210 of 214)
xxxx@xxxxxxx.... Using cached ESMTP connection to [127.0.0.1] via relay...
>>> RSET
250 2.0.0 Reset state
>>> MAIL From:<xxxx.xxxxx@xxx.xxxxxxx.xx.xx> SIZE=267525
451 4.3.2 Please try again later
xxxx@xxxxxxx.... Deferred: 451 4.3.2 Please try again later

As far as i am able to trace it down in the sendmail source, the error 
is caused at milter initialization time. Increasing the MilterLogLevel
did not provide any usefull information btw. The problem may
dissapear and reappear after a while, restarting sendmail/dccm 
makes it go away at once. 


bart
--
Bart Dumon                                   Tel +32 2 275 31 11
UNIX System Engineer                         Fax +32 2 275 37 25
Scarlet Belgium                            http://www.scarlet.be
Medialaan 50, B-1800 Vilvoorde            bart.dumon@scarlet.biz




More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.