Let me try to explain.
One week ago I've installed a new Server 2008 R2 server with the intention to make it a mailserver. On the server I installed Exchange 2010.
For one week all services where working. (inbound/outbound mail, Outlook webaccess, internal mail in the company domain)
Since yesterday after updating and restarting my server (not sure if its relevant but its info) a strange thing is happing to the Inbound SMTP mail. (for the record all other mail operations are working)
Things I've done to debug:
When I run the Inbound smtp mail check on: test exchange connectivity I get exception:
The server returned status code 421 - Service not available, closing transmission channel. The server response was: 4.4.1 Connection timed out Exception details: Message: Service not available, closing transmission channel. The server response was: 4.4.1 Connection timed out Type: System.Net.Mail.SmtpException Stack trace: at System.Net.Mail.SmtpConnection.GetConnection(ServicePoint servicePoint) at System.Net.Mail.SmtpClient.Send(MailMessage message) at Microsoft.Exchange.Tools.ExRca.Tests.SmtpMessageTest.PerformTestReally()
When I telnet from the following locations: pc in my domain, On the mailserver itself, and a pc outside of the domain I get the exact same behaviour. I telnet to : telnet mail.mydomain.com 25 telnet gives me an answer. Then I say helo: helo mydomain.com And the answer is: 421 4.4.1 Connection timed out Connection to host is lost
But now the REALY weird part. If I leave my server alone then after a few hours Inbound SMTP mail STARTS working again???!!! And from that moment on it stays working UNTIL I reboot my server again...
The time when the service starts answering smtp calls again is totaly random of course and is leaving me with my hands in my hair how on earth I'm going to debug this problem. I've googled and tried stuff for hours now so I'm desperate for more ideas how to troubleshoot this. :)
The server is not busy at all it has CPU 0% and 1,7GB memory allocated... so its doing nothing
Does anybody have point for me to investigate this weird problem?
I noticed that I had quite a few warnings in the application log related to exchange and 1 error. All of those warnings are now fixed by setting some exchange services to delayed start. Looked like the Exchange system was starting to fast.
After this I rebooted a few times again but unfortunately fixing the warnings did not fix the time out problem.
As a test I recreated the following connectors while rebooting between every recreation to test if this might help:
Receive connector (organisation hub) Default client connector for non-MAPI (server config hub) Default hub connector also configured for internet access (server config hub) Unfortunatly these actions also did not help....