none
Exchange 2010/2003 CoExistance - Mailflow issue after problem on Exchange 2003 RRS feed

  • Question

  • Current network environment: Exchange 2003/2010 Coexistence - All mail is sent and received from EXCH03
     
    Recent Network Event: I recently had several Exchange Services that would constantly restart all of a sudden. No email would go out because of this.  I did some research and found that I had corruption in my IIS Metabase so I went ahead and uninstalled/reinstalled IIS and then performed a setup.exe /reinstall option for Exchange 2003. I then reapplied and performed all updates and mail was up and going again.
     
    I currently have 3 Vitural SMTP Servers configured. The primary Virtual SMTP server is sending out mail no problem. But one of them is not sending email between servers at all. We use this V-SMTP for service accounts. But when i perform an SMTP telnet test I immediately get a queued message. I can see the sessions connected but nothing is going out from this V-SMTP server to the Exchange 2010 mailboxes.  Anyone ever see this before??

    I'm thinking two scenerios and the first is that i'm still having some kinda of IIS problem or the second that I'm having some type of issue with my Bi-Directional Routing Group connector.

    Help would be appreciated.

    Wednesday, March 7, 2012 10:03 PM

Answers

  • On Thu, 8 Mar 2012 14:58:53 +0000, ZeroBurn7 wrote:
     
    >I restarted IIS and SMTP on my Exchange 2003 Server and no dice. I still cannot send to my Exchange 2010 Server but the others V-SMTP servers are working.
     
    So have a look at the SMTPprotocol log on the 2003 server and the SMTP
    receive protocol log on the 2010 server and see why the message
    transfer is refused.
     
    Is there a firewall or SMTP proxy between the two machines? Are *ALL*
    the ESMTP keywords sent by both machines seen by the other machine?
     
    The usual susects when this happens are Cisco devices that think the
    certain ESMTP keywords are "harmful" and simply drop them or
    substitute some made-up value for them, or firewalls that aren't "wide
    open" between the two machines.
     
    Other problems are the resolution of the FQDN that the 2003 server
    sends in the EHLO command, or the name in the 220 banner sent by the
    2010 Receive Connector.
     
    ---
    Rich Matheisen
    MCSE+I, Exchange MVP
     

    --- Rich Matheisen MCSE+I, Exchange MVP
    Thursday, March 8, 2012 10:27 PM

All replies

  • Update:

    I performed a MS Exchange BPA and got two "Missing FQDN in service principal name" errors. According to this:

    The Computer account for Exchange server serverA.domain.com does not appear to contain teh fully qualified domian name of Exchange SMTP virtual server 'm1.domainx.com'. ....  The tool expected to find 'SMTPSVC/mail.domain.com' in the 'ServicePrincipalName'. Don't know if this would cause the problem that I am currently seeing but thought I would toss it into the mix.

    Wednesday, March 7, 2012 10:40 PM
  • On Wed, 7 Mar 2012 22:40:49 +0000, ZeroBurn7 wrote:
     
    >I performed a MS Exchange BPA and got two "Missing FQDN in service principal name" errors. According to this:
    >
    >The Computer account for Exchange server serverA.domain.com does not appear to contain teh fully qualified domian name of Exchange SMTP virtual server 'm1.domainx.com'. .... The tool expected to find 'SMTPSVC/mail.domain.com' in the 'ServicePrincipalName'. Don't know if this would cause the problem that I am currently seeing but thought I would toss it into the mix.
     
    Use SETSPN to add the missing SPNs to the machine account.
     
    Missing SPNs cause Kerberos authentication failures and that casues te
    mail to remain undelivered.
     
    ---
    Rich Matheisen
    MCSE+I, Exchange MVP
     

    --- Rich Matheisen MCSE+I, Exchange MVP
    Thursday, March 8, 2012 1:08 AM
  • I added the SPN would I need to restart any services on my Exchange 2003 box where I added? Or even on my Exchange 2010 server?
    Thursday, March 8, 2012 2:33 PM
  • I restarted IIS and SMTP on my Exchange 2003 Server and no dice. I still cannot send to my Exchange 2010 Server but the others V-SMTP servers are working.
    Thursday, March 8, 2012 2:58 PM
  • On Thu, 8 Mar 2012 14:33:28 +0000, ZeroBurn7 wrote:
     
    >I added the SPN would I need to restart any services on my Exchange 2003 box where I added? Or even on my Exchange 2010 server?
     
    No. The change is made to the computer object in the AD. Exchange just
    reads the information.
     
    ---
    Rich Matheisen
    MCSE+I, Exchange MVP
     

    --- Rich Matheisen MCSE+I, Exchange MVP
    Thursday, March 8, 2012 10:21 PM
  • On Thu, 8 Mar 2012 14:58:53 +0000, ZeroBurn7 wrote:
     
    >I restarted IIS and SMTP on my Exchange 2003 Server and no dice. I still cannot send to my Exchange 2010 Server but the others V-SMTP servers are working.
     
    So have a look at the SMTPprotocol log on the 2003 server and the SMTP
    receive protocol log on the 2010 server and see why the message
    transfer is refused.
     
    Is there a firewall or SMTP proxy between the two machines? Are *ALL*
    the ESMTP keywords sent by both machines seen by the other machine?
     
    The usual susects when this happens are Cisco devices that think the
    certain ESMTP keywords are "harmful" and simply drop them or
    substitute some made-up value for them, or firewalls that aren't "wide
    open" between the two machines.
     
    Other problems are the resolution of the FQDN that the 2003 server
    sends in the EHLO command, or the name in the 220 banner sent by the
    2010 Receive Connector.
     
    ---
    Rich Matheisen
    MCSE+I, Exchange MVP
     

    --- Rich Matheisen MCSE+I, Exchange MVP
    Thursday, March 8, 2012 10:27 PM