On Mon, Oct 19, 2015 at 12:03:52AM -0400, Homer Wilson Smith wrote:
> Here are the complete logs for 2 failures and 1 success.
Absence of evidence is not ironclad evidence of absence. Just
because there's no logging of a complete delivery, does not mean
that no delivery took place. All we know is that no delivery seems
to be recorded in the logs.
> Every hour on the hour aliases are upgraded and newaliases run. When
> this happens logs show a whole mess of smtp and smtpd's recording that the
> hash table has changed and that the process is restarting.
The restarts are Perfectly normal, this happens after a client has
disconnected, and before any new connection is accepted. It has
no impact on mail delivery.
> 1.) Failure, no call to qmgr. Notice table hash updates.
>
> Oct 2 09:02:00 light postfix/smtpd[1691]: connect from
> mail1.bemta8.messagelabs.com[216.82.243.205]
> Oct 2 09:02:00 light postfix/smtpd[1691]: setting up TLS connection from
> mail1.bemta8.messagelabs.com[216.82.243.205]
> Oct 2 09:02:00 light postfix/smtpd[1691]: Anonymous TLS connection
> established from mail1.bemta8.messagelabs.com[216.82.243.205]: TLSv1 with
> cipher DHE-RSA-AES256-SHA (256/256 bits)
> Oct 2 09:02:08 light postfix/smtpd[1691]: CC97480258A:
> client=mail1.bemta8.messagelabs.com[216.82.243.205]
> Oct 2 09:02:15 light postfix/smtpd[1691]: disconnect from
> mail1.bemta8.messagelabs.com[216.82.243.205]
> Oct 2 09:02:15 light postfix/smtpd[1691]: table
> hash:/etc/aliases(0,lock|fold_fix) has changed -- restarting
>
> Oct 2 09:02:08 light postfix/smtpd[1691]: CC97480258A:
> client=mail1.bemta8.messagelabs.com[216.82.243.205]
> Oct 2 09:02:08 light postgrey[28442]: CC97480258A: action=pass,
> reason=client AWL, client_name=mail1.bemta8.messagelabs.com,
> client_address=216.82.243.205, sender=ssrs_d...@corning.com,
> recipient=d...@howellplastics.com Oct 2 09:02:09 light
> postfix/cleanup[1699]: CC97480258A:
> message-id=<628169e783974a3785796a8c12bdf...@na.corning.com>
Report logging from cleanup[1699], and qmgr shortly before and after these
entries.
My best bet is lost log entries. Make sure your syslog is configured
and working correctly.
--
Viktor.
