Tracking tables on client db or no? RRS feed

  • Question

  • I have two test projects to help me learn sync framework.

    One test project has a sql ce db on the client and I used some sample code from the sync 101 samples. It created a bunch of tracking tables on the client.

    For the second test project, I used the local database cache template. It did not generate those tracking tables.

    Why? What is the difference in the two scenarios? I do not need a peer to peer sync. Does that make a difference?

    Wednesday, November 10, 2010 5:51 PM

All replies

  • On the CE database both methods use the same built in tracking. The provider used in the sync 101 also stores extra information in some tables that you have observed.

    Look at the built in features of both methods and you will find that the sync 101 sample has more built in functionality.




    Wednesday, November 10, 2010 9:06 PM
  • I also noticed using the local database cache template, i only need to instantiate the SyncAgent with the SyncContractClient and call synchronize. I don't have to setup the remote and local providers and set up the sync orchestrator. I guess this is all done for me. Also, I didn't have to creat a scope.

    Does the fact that I am trying to ship my sync through a WCF service impact this?

    Is the method I use to sync governed by my need to filter what is downloaded to the client and what is uploaded from the client?


    • Edited by Joshua Chan Wednesday, November 10, 2010 9:23 PM Additional info
    Wednesday, November 10, 2010 9:09 PM
  • I want some of my tables to maintain synchronicity of all rows, and other tables, just some rows. Should I mix and match the two methods?
    Wednesday, November 10, 2010 9:24 PM
  • I would just use one method. The method I perfer is the one in the sync 101 sample. There is a sample that synchronizes using a WCF service.



    Thursday, November 11, 2010 12:36 AM