“Merge Subscribe Once, Copy and Deploy Multiple Times” – This has been the real good feature I like and unfortunately not many people knew about it. This is applicable for v3.0 and above. Here is a short story of it:
The initial subscription is going to bring down lot of data from server to client in merge replication. It means that the requirement for bandwidth, memory, processing power is going to be high for the creation of subscription. However, for devices all these requirements are not easy and even if they are available, it takes very long time to get the subscription created on the client database. Imagine a deployment of 1000+ devices (usual number) and doing this creation of subscription for every device. This has been the major pain and customer feedback we received in v2.0 and wanted to address this in v3.0 and we didJ.
Before going into details on how it works and using it, I first want to call out the limitations/conditions with it:
C1) It is useful only in case of static filtering or no filtering. If your publication has dynamic filtering, then this does not work.
C2) Every subscribed DB copy must do an empty/dummy sync (to update the metadata both on client and server about this copy) from deployed device/SD Card/desktop before using it for local database operations/in application.
You would have loved to have:
1) Subscribe to the publication once (on the desktop! where there is required processing power, memory and optional bandwidth)
2) Copy the *same* subscribed database file to 1000+ SD Cards/Devices/Desktops
3) What else, start using it** J
It works exactly as you wantedJ. The mark ** on (3) meant to hint the condition (C2) above. Note this also means that you can deploy the database file along with your application 🙂
Namaste / Namaskaram / Thanks,
Laxmi Narsimha Rao ORUGANTI