Configure Web Synchronization
The Web synchronization option for SQL Server Merge Replication enables data replication using the HTTPS protocol over the Internet. To use Web synchronization, you first need to perform the following configuration actions:
Create new domain accounts and map SQL Server logins.
Configure the computer that is running Microsoft Internet Information Services (IIS) to synchronize subscriptions.
Configure a merge publication to allow Web synchronization.
Configure one or more subscriptions to use Web synchronization.
If you plan to replicate large volumes of data or use large data types such as varchar(max), read the section "Replicating Large Volumes of Data" in this topic.
To successfully set up Web synchronization, you must decide how you will configure security to meet your particular requirements and policies. It is best to make these decisions and create the necessary accounts before you attempt to configure IIS, the publication, and subscriptions.
In the procedures that follow, a simplified security configuration using local accounts is described, for brevity. This simplified configuration is suitable for installations where both IIS and the SQL Server Publisher and Distributor are running on the same computer, even though it is much more likely (and recommended) that you will use a multiple-server topology for a production installation. You can substitute domain accounts for the local accounts in the procedures.
Creating New Accounts and Mapping SQL Server Logins
The SQL Server Replication Listener (replisapi.dll) connects to the Publisher by impersonating the account specified for the application pool that is associated with the replication web site.
The account used for the SQL Server Replication Listener must have permissions as described in Merge Agent Security, under the section "Connect to the Publisher or Distributor." In summary, the account must:
Be a member of the Publication Access List (PAL).
Be mapped to a login associated with a user in the publication database.
Be mapped to a login associated with a user in the distribution database.
Have Read permissions on the snapshot share.
If this is the first time you are using SQL Server Replication, you will also need to create accounts and logins for the replication agents. For more information, see the "Configuring the Publication" and "Configuring the Subscription" sections in this topic.
Before you configure Web synchronization, we recommend that you read the "Security Best Practices for Web Synchronization" section in this topic. For more information about Web synchronization security, see Security Architecture for Web Synchronization.
Configuring the Computer That Is Running IIS
Web synchronization requires that you install and configure IIS. You will need the URL to the replication Web site before you can configure a publication to use Web synchronization.
Web synchronization is supported on IIS beginning with version 5.0. The Configure Web Synchronization Wizard is not supported on IIS version 7.0. Beginning with SQL Server 2012, to use the web sync component on IIS server, we recommend users install SQL Server with replication. This can be the free SQL Server Express edition.
SSL is required for Web synchronization. You will need a security certificate issued by a certification authority. For testing purposes only, you can use a self-issued security certificate.
To configure IIS for Web synchronization
SQL Server Management Studio: Configure IIS for Web Synchronization
SQL Server Management Studio: Configure IIS 7 for Web Synchronization
Creating a Web Garden
The SQL Server Replication Listener supports two concurrent synchronization operations per thread. Exceeding this limit may cause the Replication Listener to stop responding. The number of threads allocated to replisapi.dll is determined by the application pool Maximum Worker Processes property. By default, this property is set at 1.
You can support a greater number of concurrent synchronization operations per CPU by increasing the Maximum Worker Process property value. Scaling out by increasing the number of worker processes per CPU is known as creating a "Web garden."
Web gardening will allow more than two Subscribers to synchronize at the same time. It will also increase CPU utilization by replisapi.dll, which can negatively impact overall server performance. It is important to balance these considerations when you choose a value for Maximum Worker Processes.
To increase Maximum Worker Processes in IIS 7
In Internet Information Services (IIS) Manager, expand the local server node, and then click on the Application Pool node.
Select the application pool associated with the Web synchronization site, and then click Advanced Settings on the Actions pane.
On the Advanced Settings dialog, under the Process Model heading, click the row labeled Maximum Worker Processes. Change the property value, and then click OK.
Configuring the Publication
To use Web synchronization, create a publication in the same way that you would for a standard merge topology. For more information, see Publish Data and Database Objects.
After the publication is created, enable the option to allow for Web synchronization by using one of the following methods: SQL Server Management Studio, Transact-SQL, or Replication Management Objects (RMO). To enable Web synchronization, you will need to supply the Web server address for Subscriber connections.
If you are using a Publisher for the first time, you must also configure a Distributor and a snapshot share. The Merge Agent at each Subscriber must have read permissions on the snapshot share. For more information, see Configure Distribution and Secure the Snapshot Folder.
gen is a reserved word in websync xml files. Do not attempt to publish tables containing columns named gen.
Configuring the Subscription
After you enable a publication and configure IIS, create a pull subscription and specify that the pull subscription should synchronize by using IIS. (Web synchronization is supported only for pull subscriptions.)
Upgrading from an Earlier Version of SQL Server
If you have an existing Web synchronization topology configured and you upgrade SQL Server, you must ensure that the latest version of Replisapi.dll is copied to the virtual directory used by Web synchronization. By default, the latest version of Replisapi.dll is located in C:\Program Files\Microsoft SQL Server\<nnn>\COM.
Replicating Large Volumes of Data
To help avoid potential memory problems on Subscriber computers Web synchronization uses a default maximum size of 100 MB for the XML file used to transfer changes. The limit can be raised by setting the following registry key:
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Microsoft SQL Server\100\Replication
WebSyncMaxXmlSize DWORD 2000000
The range of acceptable values for this key is 100 MB to 4GB. The value is specified in KB. Setting this parameter to a high value does not guarantee that you can synchronize that amount of data. The effective limit is constrained by how much contiguous memory is available on the Subscriber computer. If you must have a value larger than 100 MB, we recommend that you increase the value incrementally and test memory consumption with a typical workload on the Subscriber.
The maximum size for the XML file is 4 GB, but replication synchronizes the changes from that file in batches. The maximum batch size of data and metadata is 25 MB. You must ensure that the data in each batch does not exceed approximately 20 MB, which allows for metadata and any other overhead. This limit has the following implications:
You cannot replicate any column that causes the data and metadata to exceed 25 MB. This might be an issue when you are replicating rows that contain large data types, such as varchar(max).
If you replicate large volumes of data, you might have to adjust the Merge Agent batch size.
Batch size for merge replication is measured in generations, which are collections of changes per article. The number of generations in a batch is specified by using the–DownloadGenerationsPerBatch and –UploadGenerationsPerBatch parameters of the Merge Agent. For more information, see Replication Merge Agent.
For large volumes of data, specify a small number for each of the batching parameters. We recommend that you start with a value of 10, and then tune based on application needs and performance. Typically, these parameters are specified in an agent profile. For more information about profiles, see Replication Agent Profiles.
Security Best Practices for Web Synchronization
There are many choices for security-related settings in Web synchronization. We recommend the following approach:
The SQL Server Distributor and Publisher can be on the same computer (a typical setup for merge replication). However, IIS should be installedon a separate computer.
Use Secure Sockets Layer (SSL) to encrypt the connection between the Subscriber and the computer running IIS. This is required for Web synchronization.
Use Basic Authentication for connections from the Subscriber to IIS. Using Basic Authentication, IIS can make connections to the Publisher/Distributor on behalf of the Subscriber without requiring delegation. Delegation is required if you use Integrated Authentication.
Basic Authentication is the method by which credentials are passed to IIS. Basic Authentication does not prevent specifying Windows domain accounts for connections that are made to IIS.
Specify that the Snapshot Agent should run under a Windows domain account, and specify that the agent should make connections as that account. (This is the default configuration.) Specify that each Merge Agent should run under the domain account of the user that uses the Subscriber computer, and specify that the agent should make connections as that account.
For more information about the permissions that are required by agents, see Replication Agent Security Model.
Specify the same domain account as the one the Merge Agent uses when you specify an account and password on the Web Server Information page of the New Subscription Wizard or when you specify values for the @internet_url and @internet_login parameters of sp_addpullsubscription_agent. This account must have read permissions for the snapshot share.
Each publication should use a separate virtual directory for IIS.
The account under which the SQL Server Replication Listener (Replisapi.dll) runs is also the account that will connect to the Publisher and Distributor during synchronization. This account must be mapped to a SQL Login account on the Publisher and Distributor. For more information, see the "Setting Permissions for the SQL Server Replication Listener" section in the Configure IIS for Web Synchronization.
You can use FTP to deliver the snapshot from the Publisher to the computer that is running IIS. The snapshot is always delivered from the computer that is running IIS to the Subscriber by using HTTPS. For more information, see Transfer Snapshots Through FTP.
If servers in the replication topology are behind a firewall, you might need to open ports in the firewall to enable Web synchronization.
The Subscriber computer connects to the computer that is running IIS over HTTPS using SSL, which is typically configured to use port 443. SQL Server Compact Subscribers can also connect over HTTP, which is typically configured to use port 80.
The computer that is running IIS typically connects to the Publisher or Distributor using port 1433 (default instance). When the Publisher or Distributor is a named instance on a server with another default instance, port 1500 is typically used to connect to the named instance.
If the computer running IIS is separated from the Distributor by a firewall and an FTP share is used for snapshot delivery, the ports used for FTP must be opened. For more information, see Transfer Snapshots Through FTP.
Opening ports in your firewall can leave your server exposed to malicious attacks. Make sure that you understand firewall systems before you open ports. For more information, see Security Considerations for a SQL Server Installation.