2003 to 2010 Public folders not moving replicas

[Head-Hunter]

Hi guys. I have finished the migration of Exchange 2003 to 2010 wihtout any issues - I moved through the applicable suggestions that I found (MS one, tons of blog ones etc). All of things are working fine as it should - I have a mix of 2003, 2007 and 2010 Outlook clients, and all of them running without any problems with the Emails, calendars, public directories, free/busy etc. To replicate Public directories , first I used AddReplicaToPFRecursive.ps1 script to replicate them to 2010 Exchange. After some days and all of thing functioning fine , I used MoveAllReplicas.ps1 to initiate move process. After some days, I still get Public directory instances on 2003 populated with entire directories (however they were moved from Public directory container in ESM - they are implemented to be there in the same way). If I click on the 2003 ESM within the Public directories, I only get two directories at there: Default and exchangev1.

In 2003 Application log, I got the tons of EventID 3093

"Event Type: Error
Event Source: MSExchangeIS Public Store
Event Category: Replication Errors
Event ID: 3093
Date: 7/23/2010
Time: 2:09:00 PM
User: N/A
Computer: LAEXCHANGE1
Description:
Error -2147221233 reading property 0x67480014 on object type tbtOwningFolders from database "First Storage Group\Public Folder Store (LAEXCHANGE1)".

[Ekanga]

If you choose "Remove Replica" making a right click within the ESM, then this will clean out the replica, but would not delete the entry Public Folder Instances. That will not delete out until entire things is replicated and copied upon without any issue and any type of error message . (you have actually already finished the same when you go for moving entire replicas command.

If you connect with the 2010 PF version of the calendar via ExFolders, have you enough permissions? If you are unable to get rid of such sections , then you may have to end up in the manual way to remove out the 2003 PF database from the ADSI.

[Head-Hunter]

I thought that i can solve it myself and I started seeking some more about this , and when I concentrated at my Queues for 2003 LAEXCHANGE1 (mailbox server where public directories exits ), I got that the specified 2010-2003 connector has some thousand messages within this. Somehow replication messages are not routed in proper way with my CAS 2010 server. If any suggestions what could be the issue there?

[Ekanga]

Initially, Email flows to the HUB server and not the CAS. Are you getting that is there some Email flowing among two servers? One for thing to find out here...is the Exchange 2003 server another part of the ExchangeLegacyInterop security section ? If not it requires to be.

Another step is to remake the routing section connectors. From the 2003 server, ESM remove the two different instances of the routing group connector.you have to go first on to the 2010 section, from EMS run:

New-RoutingGroupConnector -Name "Interop RGC" -SourceTransportServers "Ex2010Hub1.contoso.com" -TargetTransportServers "Ex2003BH1.contoso.com" -Cost 10 -Bidirectional $true -PublicFolderReferralsEnabled $true

[Head-Hunter]

Thanks for the suggestion . I have again created the connectors, and few of the thousands of messages got stuck within the queue moved through. After some minutes, I got to see that my Public directory instances moved down to three directories: exchangeV1, one of the public directories (shared calendar), and Shedule Free/busy (First Admin section one). Almost there ..

As far as connector moves: I need to be sure that the actual servers are connected and my topology is described below:

Exchange2003 Front-End and Exchange2003 Back-End (mailbox/PF server) connected with RPC-HTTP.

Exchange2010 Front (CAS role only), and Exchange2010 Back (HUB/MBX roles).