These are two separate issues, both having to do with frozen messages.
The Mail queue is still freezing messages upon arrival. I have to run a
manual exim -qff to make it unfreeze and deliver everything in the
queue. It's really getting annoying. I finally just made a crontab
entry to do it for me.
However, _A_LOT_ of the messages that ware frozen were because of the
ECN stuff. BTW the reason I rebooted is because I managed to break
something in my config while I was at it, and rebooting was
faster/easier than correcting the problem.
Could the system load average have something to do with why exim is
freezing messages rather than delivering them? What is the default
behaviour for that feature? As I've said before, the docs online are
about as helpful as a kick in the pants, even _IF_ you do know something
about how internet e-mail works.... The instructions on what options do
what are still annoyingly vague.
I do however, wish that they had a separate mailing list for the 4.0
users since they're so different... I tried updating to 3.35 to see if
it fixed my problems, but I couldn't get one to compile and work with my
system and I really don't like dropping pre-compiled binaries into my
setup that aren't from Mandrake directly. That plus for some reason
every time I try to install something that's not from an RPM I wind up
hosing my server. ;-) Not a good thing.
> -----Original Message-----
> From: exim-users-admin@???
> [mailto:exim-users-admin@exim.org] On Behalf Of Derrick 'dman' Hudson
> Sent: Thursday, June 13, 2002 10:24 PM
> To: 'Exim-User (E-mail)'
> Subject: [Exim] Re: Re: Frozen incoming messages?
>
>
> --
> On Thu, Jun 13, 2002 at 09:58:33PM -0500, Cory Daehn wrote:
>
> | By the way, I finally got a response on a good bit of my frozen
> | messages.
> |
> | Sure enough, in /etc/sysctl.conf I had:
> | net.ipv4.tcp_ecn = 1
>
> Oh, ECN. Stupid broken routers. (As I understand it, ECN is
> backwards-compatible with the TCP *standards*, but a number
> of implementations can't deal with it.)
>
> | Just set that to off and rebooted...
>
> You REBOOTED!? This isn't windows. Just run
> /sbin/sysctl -p
> after editing /etc/sysctl.conf, or merely run
> echo 0 > /proc/sys/net/ipv4/tcp_ecn
> (which won't persist through a reboot, btw)
>
> | Problem solved with connections timed out to a lot of servers.
>
> Now I wonder why the error message said it couldn't get root
> to setuid, but the problem was a TCP timeout. Were all the
> error messages the same?
>
> -D
>
> --
>
> (A)bort, (R)etry, (T)ake down entire network?
>
http://dman.ddts.net/~dman/
--
[ Content of type application/pgp-signature deleted ]
--
--
## List details at
http://www.exim.org/mailman/listinfo/exim-users Exim
details at
http://www.exim.org/ ##
