Exchange 2003 sending email times out around 300k
I have a mailserver, Enterprise Exchange 2003, sp2 that has trouble sending to a few domains. Email gets stuck in queue for retry over and over... to isolate this I created a connector for "problem" domains and specified them, sending this mail through a smarthost that I can control and monitor logs on (an mdaemon server). I can view the logs on the mdaemon server and see a connection from the exchange box, and around 300k gets transferred within 4 to 5 minutes, then there is a connection timeout that occurs. The exchange server is on a 3x3 synchronous connection, the mdaemon server is on a 15x15 synchronous. On the exchange network side there is a barracuda spam filter (for incoming mail only) and a barracuda web appliance (with exclusions for the exchange server). The default gateway for the exchange server is the internet router; no throttling or firewalling done on outgoing traffic on it. Any ideas would be welcome, this has been going on for awhile and we don't seem to be making any progress here. TIA!!
ASKER
Thank you for your suggestions!
I will try network monitoring next.
There is another exchange server in the domain, and I've redirected mail through it, and we experience the same problem symptoms. I've deleted these messages and recreated large email and it continues to occur. One of the servers has AV scanning software running, the other does not; there is no difference in behavior.
Thanks!
I will try network monitoring next.
There is another exchange server in the domain, and I've redirected mail through it, and we experience the same problem symptoms. I've deleted these messages and recreated large email and it continues to occur. One of the servers has AV scanning software running, the other does not; there is no difference in behavior.
Thanks!
ASKER
It looks like I have this narrowed down to companies that are using an email filter appliance or third party and large attachments. You can see the screen shot attached, this is message tracking on the exchange 2003 side for one message; these are common occurrences I see with Postini and Barracudas.
There is a very large organization that has recently directed all of their incoming email scanning to Postini, and the recipients have little to no control over the settings. Does this screenshot look like something you guys have come up against before?
postiniedit.jpg
There is a very large organization that has recently directed all of their incoming email scanning to Postini, and the recipients have little to no control over the settings. Does this screenshot look like something you guys have come up against before?
postiniedit.jpg
yeah, Routers Like Cisco PIX have their own e-mail screening features. what you can do is to check the detail settings on the router/appliance and disable all e-mail related features.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
a. Enable network monitor, take a capture of the server communication, you need to find out at which point the error happens.
b. On Exchange side, you can copy the queued mails (eml files from the queue folder) out. Put them into a test clean server to reproduce this issue. If the problem still happens, this will help you narrow down the root cause to the corrupt mail items.