That's actually two sides of a single bidirectional RGC. It was created by the 
installation process and IF you provided accurate info during the installation 
process (i.e., you identified the proper legacy bridgehead and the proper new 
bridgehead) it should be fine.

Regards,

Michael B. Smith
Consultant and Exchange MVP
http://TheEssentialExchange.com

From: Paul Steele [mailto:[email protected]]
Sent: Wednesday, May 05, 2010 11:01 AM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

Thanks for the info. I'll have to take some time to digest your points. One 
thing that jumps out though is your comment that there should only be one 
Routing Group Connector. On the Exchange 2003 server, there are actually two 
RGCs, one called EXCH2003-EXCH2010 in the First Admin Group, and another called 
EXCH2010-EX2003 in the newly created Exchange Admin Group. These have not been 
modified, in fact ESM doesn't let you, complaining about version issues. Are 
these the RGs that you're referring to? I assumed they are both needed since 
they were created as part of the install process.

From: Michael B. Smith [mailto:[email protected]]
Sent: May-05-10 11:25 AM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

I would need to see email headers, and/or smtp logs, to determine how email is 
actually currently flowing.

It sounds to me as if either the RGC is incorrectly configured or you have the 
send-connector on the 2010 side misconfigured.

Actual attempted email flow (message headers or smtp logs) from the 2003 -> 
2010 and vice versa would identify the guilty party.

Regards,

Michael B. Smith
Consultant and Exchange MVP
http://TheEssentialExchange.com

From: Paul Steele [mailto:[email protected]]
Sent: Tuesday, May 04, 2010 5:27 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

Ok, I read your Windows IT Pro article, which was very good. Unfortunately it 
didn't help and I'm officially *very* frustrated now. After fiddling with 
send/receive connectors for last couple of days, I still haven't gotten mail to 
flow within our new Exchange 2003/2010 organization properly. My ultimate goal 
is to have mail arrive on Exchange 2010 and if necessary route mail to Exchange 
2003 for mailboxes that are still on the 2003 server. At this point no mail 
boxes have been moved, although I've created a couple of test 2010 accounts. I 
can send mail from 2003 users to 2010 mailboxes, and I can send mail externally 
directly to 2003 mailboxes or 2010 mailboxes.  Unfortunately I cannot get mail 
working from 2010 mailboxes to 2003 mailboxes, or externally through the 2010 
server to 2003 mailboxes. This issue is also causing Public Folder Replication 
problems.

After the 2010 install, two routing group connectors were created 
(Exch2003-Exch2010 in the First Administrative Group side, and 
Exch2010-Exch2003 in the newly created Exchange Admin Group). I haven't 
modified those in any way. I also had an existing SMTP Connector with a Unix 
smart host specified. The SMTP address space is * and applies to the entire 
organization. This SMTP connector also appears on the 2010 server as a Send 
Connector. Finally, there is an SMTP Virtual Server under the SMTP protocol 
section of the First Admin Group, with no smart host specified.

After the installation, I created scoped Send Connector for SMTP/*, specifying 
our Unix smart host and the 2010 server as the source server. Under Accepted 
Domains, I added acadiau.ca as an Internal Relay. There is also another SMTP 
Virtual Host in the Exchange Admin Group.

I actually had this all working in my ESXi test environment but due to some 
disk issues I lost my VMs. I thought I remembered how I had configured things 
but obviously I don't have a very good memory. I've gotten totally confused 
about what send connectors, SMTP virtual hosts, smart hosts, accepted domains, 
etc, etc, that I need. I think I've tried every obvious configuration. I know 
it should be simple, but it seems to have beaten me.

What am I missing?

From: Paul Steele [mailto:[email protected]]
Sent: May-03-10 2:52 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

Mail delivery seems to be working properly now, so I think I've got the various 
connectors configured properly. The 2003 server does not have a certificate and 
I haven't installed a UC cert on 2010 yet. I'm reading everything I can find on 
the topic before tackling the certificate install. After the way things went 
today I doubt that things will go problem free...

From: Michael B. Smith [mailto:[email protected]]
Sent: May-03-10 2:09 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

How have you configured namespace sharing, i.e., namespaces vs. SSL 
certificates on the two servers?

Regards,

Michael B. Smith
Consultant and Exchange MVP
http://TheEssentialExchange.com

From: Paul Steele [mailto:[email protected]]
Sent: Monday, May 03, 2010 1:07 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

That might be it. I had sort of wanted to keep Exchange 2010 out of the picture 
after this first phase since all the mail boxes are still on Exchange 2003. I 
want to get backups, certificates, and other misc stuff working on Exchange 
2010 before moving mailboxes.

One weird problem that has surfaced since the 2010 install is a problem with 
Public folders. When I try to look at the folder in System Manager on Exch2003, 
I get the error

"The HTTP service used by Public Folders is not available, possible causes are 
that Public stores are not mounted and the Information Store service is not 
running."

I'm getting lots of google hits for this error but none specifically relating 
to an Exchange 2010 upgrade. The folders show up fine in Outlook, but not in 
OWA or System Manager. This issue did not come up in our test lab.

From: Michael B. Smith [mailto:[email protected]]
Sent: May-03-10 1:49 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

That would tend to indicate that the mail is being routed through the Exchange 
2003 server and you don't have  a properly scoped send-connector on the 2010 
side.

Regards,

Michael B. Smith
Consultant and Exchange MVP
http://TheEssentialExchange.com

From: Paul Steele [mailto:[email protected]]
Sent: Monday, May 03, 2010 12:47 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

Yes, I had already done that. I'm a bit confused now though. I stopped/started 
the SMTP virtual host (on the EXCH2003 side), and that seemed to solve the 
problem. Things seem to be working so I guess whatever was broken/stuck is ok. 
At least it's working...

From: Michael B. Smith [mailto:[email protected]]
Sent: May-03-10 12:40 PM
To: MS-Exchange Admin Issues
Subject: RE: Exchange 2003 to 2010 issues

Did you create a send-connector to forward the scoped domain to the Unix host?

Regards,

Michael B. Smith
Consultant and Exchange MVP
http://TheEssentialExchange.com

From: Paul Steele [mailto:[email protected]]
Sent: Monday, May 03, 2010 11:37 AM
To: MS-Exchange Admin Issues
Subject: Exchange 2003 to 2010 issues

We did as much testing as we could in our test lab, so we moved ahead and 
installed our production Exchange 2010 server. The installation went fine 
without any errors, but we're have delivery problems.

We've been preparing for a transition from Exchange 2003 to Exchange 2010. 
We've done various test installs in our  lab and thought we had everything 
figured out. Unfortunately we're having mail delivery  problems after 
completing the install. The actual installation went without problems. We have 
a shared address space, with some users on our Exchange server and some on a 
Unix host. The biggest issue we're having is sending to users who are not on 
our Exchange server. We've set the Accepted Domains for "acadiau.ca" to 
Internal Relay, but when we try to send to 
[email protected]<mailto:[email protected]> that does not exist in the 
Exchange name space, the message bounces (it never makes it to our Unix smart 
host). We've compared things with our test servers and everything looks 
correct. At this point we've very confused what we're doing wrong. If anyone 
has a suggestion what to look for we really appreciate it.

Reply via email to