Papo Napolitano writes:
I'm observing a rather extrange behaviour (or maybe a configuration error) in the latest courier (0.52.2) build.I'm filtering using maildrop and a custom perl filter which right now it's taking a lot more than 5 minutes (the default watchdog), so I put "30m" in my localtimeout file as man courier says: localtimeout This file specifies the watchdog timer for local mail deliveries. If a local mail delivery attempt does not complete in the proscribed time interval, the delivering process ID is killed. The time interval in localtimeout is specified in the same way as queuetime (see queuetime for more information). Well, this isn't happening, as I see this on the logs: Jan 17 13:14:41 localhost courierd: started, id=006B85BB.43CC1C7B.00007CCA, from=<[EMAIL PROTECTED]>, module=local, [EMAIL PROTECTED]/home/mail/dosalcubo.com/test!!, addr=<[EMAIL PROTECTED]> Jan 17 13:19:41 localhost courierlocal: id=006B85BB.43CC1C7B.00007CCA, from=<[EMAIL PROTECTED]>, addr=<[EMAIL PROTECTED]>: maildrop: Timeout quota exceeded.
localtimeout is Courier's local mail delivery timeout.You are running maildrop as part of local mail delivery, and maildrop has its own 5 minute, fixed timeout.
Mail should not take more than five minutes to get delivered. If it's taking longer than that, something is broken.
The number of local mail delivery slots is limited. A broken local mail delivery script will quickly use up all available delivery slots and block delivery of mail to other mailboxes. This is unacceptable.
You're trying to fix the symptom, instead of the underlying disease. Fix the disease instead.
pgpnC4Iacp3RC.pgp
Description: PGP signature
