On Fri, Feb 18, 2011 at 09:42:54AM +0000, Maria Arrea wrote:
> Hello
>
> I have Red Hat Enterprise Linux 5, postfix 2.3.3 and dovecot
> 2.0.9. Mails are delivered to dovecot by virtual_transport=dovecot and
> the mailbox_command is /usr/libexec/dovecot/deliver. Everything works
> as expected.
Of these two settings may be unnecessary. Is the domain in question a
virtual mailbox domain or a local domain (listed in mydestination)? Or
do you have some of each and deliver all to dovecot? If so, generally
one would usually use mailbox_transport, not mailbox_command, otherwise
the dovecot mailbox ownership would be inconsistent...
> My first question is about logs. Without compression in dovecot I got log
> lines like this:
>
> /Feb 4 14:10:13 buzon postfix/pipe[4018]: 4EA77C7775: to=<cor...@us.es>,
> relay=dovecot, delay=1.6, delays=1.3/0.17/0/0.12, dsn=2.0.0, status=sent
> (delivered via dovecot service)/
Averages are much more useful than single entries. Compute and report ~N=10
sample averages, each taken over 1/Nth of the total deliveries.
> Feb 3 17:48:25 buzon postfix/pipe[8263]: 2169CC7758:
> to=<javierdemig...@us.es>, relay=dovecot, delay= *13*, delays=
> *0.83/11/0/1.2*, dsn=2.0.0, status=sent (delivered via dovecot service)
What is the CPU load at this time?
> With compression in mailboxes: delay= *13*, delays= *0.83/11/0/1.2*
> Without compression: /delay=1.6, delays=1.3/0.17/0/0.12
>
> The big difference is the second value, /"b - time in queue manager"/. When
> is the mailbox_command executed & measured ? In b or in d ?
The "b" value is a symptom, not a cause. Because delivery is slow, mail
that arrives later waits in the queue.
> My second question: I use syslog for logging. I have the following
> line in /etc/syslog.conf
>
> mail.* -/var/log/maillog
>
> What can I do to improve logging perfomance? I would like to stick with
> standard syslog.
You don't need to improve logging performance, it works fine when mail
is not compressed and disk I/O loads are presumably higher.
--
Viktor.
