Improve synchronization performance between non-cloud databases and Windows Azure SQL databases

We are pleased to announce that the SQL Data Synchronization service has been updated and can now operate in all Windows http://www.aliyun.com/zixun/aggregation/13357.html >azure data centers. The past two updates (June, August), in addition to the general reliability improvements for the preview version, bring the following enhancements:

• Enhanced support for spatial data types, including geography and geometry
• New features: Users can cancel ongoing data synchronization
• Improved overall performance of initial provisioning and synchronization tasks
• Improved synchronization performance between non-cloud databases and Windows Azure SQL databases
SQL data synchronization enables us to create and plan for regular synchronization between Windows Azure SQL databases and SQL Server or SQL databases. You can get more information about SQL data synchronization from MSDN. We have also published SQL data synchronization best practices on MSDN.

Our team is working hard for subsequent updates, moving towards universal applicability, and we sincerely thank you for your timely feedback. Please keep feedback and use Windows Azure SQL Database forum to ask questions and ask for assistance. Do you have a particular favorite feature in SQL data synchronization? Use the "Functional polling forum" to cast a vote for the features you expect to add or upgrade.

The following are instructions for synchronizing SQL data:

This article is intended for all users who plan to use SQL data synchronization, especially at the planning and design stages. This article includes the following topics: Select a location for the SQL database instance, make sure that the SQL database instance you create has enough space to handle data and SQL data synchronization, select the tables to include in a particular synchronization group, and so on.

The SQL Data Synchronization service allows bidirectional data synchronization between two or more geographically dispersed SQL database instances and SQL Server databases. SQL Database instances can exist in any of the Windows Azure data centers around the world. SQL Server databases can also be geographically dispersed. You must have at least one SQL database instance for SQL data synchronization to use as a central database for synchronization.

Security and reliability

Agent

Install the client proxy using the least privileged account with access to the network service. It is a good idea to install the client agent on a separate computer that is not your local SQL Server computer. Do not register the local database with multiple agents.





even synchronizing different tables in different synchronization groups. Registering a local database with multiple client agents can cause problems when a synchronization group is deleted.

Database account with least privilege

for sync Group:


Create/change tables, change databases, create procedures, select/Change schemas, create user-defined types. For synchronizing in progress:


performs select/insert/update/delete operations on the tables selected for synchronization, synchronization metadata, and trace tables, execute permissions on stored procedures created by our services, and execute permissions on user-defined table types. For cancellation:


changes to tables that are synchronized, control of synchronization meta 6184.html > datasheet, synchronization Tracking tables, stored procedures, and user-defined types.

How do I handle information that results from having only one credential for a database in a synchronization group?

changes the credentials for the different stages (for example, cred1 for the installation phase, CRED2 for the ongoing phase). Change permissions for credentials (for example, change permissions after installation synchronization).

Placement location of the central database

Enterprise to Cloud Scenario:

Keep your central database close to the center of the synchronization group's database data traffic to minimize latency.

Cloud to Cloud Scenario:

When all the databases in a synchronization group are located in one datacenter, the central database should be located in the same data center. This reduces latency and the cost of transferring data between data centers.

When a database in a synchronization group is located in multiple data centers, the central database should be located in the datacenter where most database and database data traffic resides.

Mixed solution:

Apply these principles to more complex synchronization group configurations.