Zen Internet Support Forum

3 times in the past 2 weeks, my exchange 2003 server outgoing smtp queue has got blocked with stuff not going.

Servers like hotmail.co.uk and then others, not quite so widely known have all been queuing up.

 At first I thought it might be my exchange server (heck - always blame MS first! :)) - so I looked up the MX record for one of the queues and telneted in on port 25 to try send an email by hand - but the connection kept getting dropped or the remote server reported bad syntax (I am pretty sure that my SMTP syntax was fine - I was using a prewritten text file to paste in - and sometimes it worked - other times not).

 I switched my router / firewall on & off and once I'd reconnected all was well.

All other internet access was running fine (including VPNS to other servers and FTP sessions) - inbound SMTP was running just fine.

No configuration changes have been made to the firewall in years.

Any suggestions as to the cause? The router/firewall is a Vigor2600+ and I'm syncing at 6752000 down and 800000 up - S/N  margin at 6.5 and loop attenuation at 37.5 - which is pretty much what I've had for ages.

 Thanks.

M

Bit of a long shot perhaps but I have some Vigor 2600 routers which from time to time return the bytes of the IP address of external mail servers in reverse order. ie mail.ourdomain.com should be 1.2.3.4 but when we ping it from a machine it tries to connect to 4.3.2.1 - only seems to do it with IP addresses which are accessed a lot.

Thanks for the reply. I don't think that is the problem as the router doesn't do any DNS lookups or anything - and the problem also occured when I tried to send an email by hand using telnet.

All very odd. It has since happened 3 more times.

Try this:

 telnet <mailserver> 25
helo <yoursmtpserverIP>
mail from: <youremailaddress>
rcpt to: <recipientemailaddress>
data
subject: Test
.

If this works restart the System Attendant service and check your DNS.

Yes - that is the thing. That is what I meant by "sending email by hand". When in an error state - doing that returns syntax or other errors from the host server - as if port 25 traffic is getting garbled. It is still going on, and a reboot of the router solves the problem though.

I have tried telneting from other machines on my network - all with the same result when the exchange server is in error state- and all suffer the same problem. This implies it isn't a fault with the exchange server, but either the router or the connection. This is still very puzzeling.

 DNS works fine - I have no problems looking up other servers etc.

M

I give up on this one.

I've not been able to find a cure, or even an explanation. It doesn't look like it is the exchange server, and I can't see why the router would suddenly start doing this after many years of service (why just garble traffic on port 25?).

So I've written an app that monitors the size of the exchange queue and if it detects a fault reboots the router for me.