Peer-2-Peer Table Mapping -- Is it *possible*??

• Question

• 

  

  0

  Sign in to vote

  I have a working peer-2-peer sync app that syncs several tables between 2 SQL Server databases (both on 2008 server). One table is called Countries and has only a few columns. I sync this between my 2 peers setting up the providers with the following code:

   

  DbSyncTableDescription CountriesTableDescription = SqlSyncDescriptionBuilder.GetDescriptionForTable( "Countries", (System.Data.SqlClient.SqlConnection)provider.Connection );

  scopeDesc.Tables.Add( CountriesTableDescription );

   

  The rest of the code is right from the sample app. Everything works find and these tables synchronize changes as expected.

   

   

  Now I simply create a new table called Countries_Version on peer1 which is an *exact*t copy of the original Countries table.

  I want to now sync Countries_Version on peer1 with Countries on peer2; I alter the code as follows when configuring my

  providers:

   

  DbSyncTableDescription CountriesTableDescription;

  if (provider.Connection.Database.ToLower() == "peer1")

  {

      CountriesTableDescription = SqlSyncDescriptionBuilder.GetDescriptionForTable( "Countries_Version", (System.Data.SqlClient.SqlConnection)provider.Connection );

      // map to the Countries table on peer2

      CountriesTableDescription.GlobalName = "Countries";

   

  }

  else

  {

      CountriesTableDescription = SqlSyncDescriptionBuilder.GetDescriptionForTable( "Countries", (System.Data.SqlClient.SqlConnection)provider.Connection );

      // map to the Countries_Version table on peer2

      CountriesTableDescription.GlobalName = "Countries_Version";

  }

   

  Note: I use one method to configure my providers and pass “peer1” or “peer2” along with the additional

  database connection parameters required. Countries has a single column PK on a bigit identity type and Countries_Version

  has the identical set-up. Both tables have an identical set of additional columns – both tables identical other than the table names.

   

   

  When attempting to first synchronize the two databases , on the call:

   

  SyncOperationStatistics stats = orchestrator.Synchronize();

   

  I get the following error:

  Microsoft.Synchronization.Data.DbSyncException: Cannot apply changes because the local provider does not have adapters configured for the following tables that were received from the remote provider: Countries. Ensure that the correct adapters have been added to both providers for scope ‘test_scope’ and that any table mapping has been properly configured.

   

   

  And, of course, the synchronization fails.

   

   

  I am trying to create the simplest demo imaginable to demonstrate table mapping between 2 differently named tables for a sync operation.  An MS developer has indicated to me previously that all I need to do is set the GlobalName on the DbSyncTableDescription object.  I think I am doing that, as demonstrated in the above code.  But I just can’t seem to get any table mapping sync to work.

   

  Can anyone shed some light on what I am doing wrong here??  I’ve been tinkering with this for hours now, and it seems that it just shouldn’t be this difficult to map 2 tables in a synchronization. I’ve scoured the MS ADO.NET Sync 2.0 documentation, but it is virtually devoid of any real help on how to map 2 tables or shed any light whatsoever on how to use GlobalName or LocalName properties.

   

  Thanks,

  Glenn

  Monday, May 17, 2010 4:10 PM