On Sun, 2006-06-04 at 19:39 +0200, Ambrogio wrote: > Il giorno dom, 04/06/2006 alle 22.01 +0930, Tim ha scritto: > > On Sun, 2006-06-04 at 12:38 +0200, Ambrogio wrote: > > > I noted that sometime messages still in queue, whit a deferred time. > > > > I think you'd need to provide some examples/logs for someone else to > > figure that out, but a few things spring to mind: > > > > On a *network* messages could get deferred if they couldn't be delivered > > to another computer because it wasn't reachable at the time, or it > > specifically responded that it couldn't accept delivery. Or, if a DNS > > server doesn't provide an answer for the domain attempting to be > > e-mailed, at the time. > > Well, this is my first problem, but this is not the real problem. > If I send mail to redhat.com recipient, and the redhat server tell to my > sendmail to try later because he is busy, I can wait for at least 20 > minutes and retry. This job is done by sendmail (I think and I hope > well). > > The other problem I think is related to my MTA. > Messages fetched by fetchmail and processed by procmail, filtered for > spam, and routed to my local account remain queued for minutes. > When the queue reach about 20-25 messages, fetchmail pause to retrieve > mail until the queue goes below 20 messages. Do your messages eventually get through? Can you post some sample entries from /var/log/maillog regarding sendmail failing to deliver something straight away? Please do not obfuscate hostnames, domain names, or IP addresses as they could be relevant. Paul.
