MySQL Student:
> Hi,
>
> I have a postfix mail server that successfully delivers thousands of
> messages per day, primarily to the same smarthost connected to it.
> However, I have a number of messages that are stuck in the queue with
> timeout errors:
>
> said: 421 smtp01.example.com Error: timeout exceeded (in reply to end
> of DATA command))
Unfortunately, you deleted lots of useful information from the
logging, including the break-down of handshake delays and of
transmission delays.
When the receiver times out receiving data, the problem is more
likely to be at the TCP level, so no amount of verbose Postfix
logging is going to help.
Instead, consider diagnosing this with tcpdump. Common TCP-level
problems are:
- Broken TCP window scaling support in firewall or end-host. Postfix
2.6 and later have a main.cf:tcp_windowsize parameter. Setting
this to some value under 65536 will work around this, otherwise
you'd have to poke the kernel with sysctl or whatever.
- Broken IP path MTU discovery. The symptom is that TCP packets
larger than some size never get acknowledged.
- Other TCP-level breakage that only shows up in packet-level
recordings.
> I've read through the debugging and logging pages at postfix.org, and
> have increased the logging, but once you increase it to a certain
> point, it becomes impossible to separate the signal from the noise.
> How can I increase (or capture) the logging for either specific
> messages, or increase it only for messages stuck in the queue?
debug_peer_list can turn on logging for specific sites, not messages.
Wietse
