I'm right in the middle of an Exchange 2003 to 2010 transition, everything has gone fine up to this point and we are happily migrating users over to the new system. Today I added a second server running the CAS, Hub and mailbox roles, the problem I have is that straight after install the server joined the CAS array and started receiving mail, we then started getting calls to say mail was disappearing, I checked on the second server and the mail is sitting there unable to be routed through the connectors on the first server, on the new server the mail is queued in the connector but if you look at the first server there is no mail in the queue, this is for both external mail and through the routing group connector to the old system...any ideas? have to say I'm stumped on this one.....

Hi

what is the error in the queue? DNS error or delayed?

Is your second server part of your send connector to route mail out or is it setup to route to the first server?

if you create another routing group connector for the new CAS/HUB server to exchange 2003 what happens?

New-RoutingGroupConnector -Name Exchange Connector -SourceTransportServers Ex2010Hub2.domain.com -TargetTransportServers Ex2003.domain.com -Cost 10 -Bidirectional $true -PublicFolderReferralsEnabled $true

Looking at the queues we have 2 external smarthost connectors showing error 421 4 2 1 unable to connect and one saying 451.5.7.3 cant achieve exchange server authentication on the routing group connector, pretty sure that the routing group connector between the 2010 and 2003 systems is blocking the second server (first 2010 server is routing fine) would I have to add the second server into the connector? or create it's own (would this cause any issues?) I was under the impression that the first server would route the mail for the second? if not it may explain the external mail from the second server stuck in the queue as if it's going out using a different IP it would be blocked on the firewall (first server goes out on a different IP)

Right fingers crossed I think I've sorted it, there were two problems
firstly, there was a relay receive connector set up and the range overlapped the
second servers IP so it looks as though the second server was trying to use that
connector and of course it couldn't authenticate, secondly the external
connectors route through a smart host which is set to only accept traffic from
the external IP that was natted for the first server so when I added the second
in as a source server it couldn't connect, removed it as a source server and let
the first do the routing and the queues disappeared (all except shadow
redundancy that is)

• Marked as answer by carn38 2 hours 32 minutes ago