On Oct 10, 2:08, John Lawson wrote:
On Fri, 10 Oct 2003, Pete Turnbull wrote:
> On Oct 9, 22:24, John Lawson wrote:
> >
> > I am moderately curious why it is taking one to five (or more)
When did you
post this one? The headers show when it passe through
each MTA along the way (in reverse order):
This one 'cleared' fairly quickly, as you point out. Vide my
previous
message "Scrapping Tape Drive" for an
example of a 'delayed' post.
Well, I no longer have that particular message...
What I am still asking is why it sometimes takes
hours for a post
to
propagate - not how to read / interpret routing info.
I don't care
much
about the data per se, just trying to understand the
*algorithm*.
My point is that unless you can see *where* the delay is, you won't be
able to guess *why*. If a server is busy, it might queue mail up; mail
queues are normally run at regular intervals but the size of the
interval depends on the sysadmin who set it up (15 minutes is common).
If there's a temporary DNS failure, mail might be queued for longer.
There are various other things that might delay mail -- on a busy
server, set up to use idents, mail from a PC might get delayed long
enough to be deferred until the next queue run (and then deferred
again, and...)
Next time you see a delay, look at the headers in that message (or send
them to me -- I tend to delete mail as I read it, so telling me which
message, after the event, isn't going to be helpful :-))
--
Pete Peter Turnbull
Network Manager
University of York