Replication Server interacts with other components of a replication system as either a server or a client.
As a server, Replication Server supports connections from:
Replication Agents, across which database commands are sent from primary databases
Other Replication Servers, thus distributing the processing involved in message delivery and providing a degree of scalability in a replication system
Users or management tools for administration, data server identification, message publication and subscription, and so on
As a client, Replication Server connects to:
Replicate Replication Servers for message delivery
Other Replication Servers, thus distributing the processing involved in message delivery and providing a degree of scalability in a replication system
Replication Server also connects to an external Adaptive Server Enterprise database that acts as the Replication Server System Database (RSSD), if it is not configured to use an embedded RSSD.
Replication Server is an Open Client and Open Server application that uses Sybase Tabular Data Stream™ (TDS) as the underlying communication protocol. As such, any clients that request services from Replication Server must implement an Open Client interface. This includes Replication Agents, system management tools, and user interface tools such as isql.
As a client distributing messages to other Replication Servers or to replicate data servers, Replication Server speaks with an Open Client interface. Therefore, when Replication Server needs to send a message to a data server, either that data server must support an Open Server interface running on TDS, or there must be an Open Server/TDS bridge or “gateway” application between Replication Server and the replicate data server.
Typically, this gateway software is in the form of a Sybase DirectConnect database gateway. Some DirectConnect gateways bridge from Open Server/TDS to the native interface of the replicate data server (for example, DirectConnect for Oracle), while others bridge from Open Server/TDS to an ODBC or JDBC driver for the data server. Replication Server configurations vary, depending on the gateway used.
Replication Server requires several different user IDs. Some user IDs are required for other components (or users) to access the Replication Server, and others are required for the Replication Server to have access to other components in a replication system.
User IDs are defined in the Replication Server using the Replication Server create connection command.
Depending on how your replication system is configured,
some of the user IDs in the following list might not be required.
For example, if you have separate Replication Servers for primary
and replicate databases, the primary Replication Server does not
require a user ID for access to a replicate database.
The following user IDs are defined in a Replication Server:
Replication Agent user – used by a Replication Agent to log in to a primary Replication Server. This user ID must have connect source permission to deliver database commands through the LTL interface.
Replication Server user – used by other Replication Servers to log in to a Replication Server and forward messages. This user ID must have connect source permission to forward database commands through the RCL interface.
SysAdmin user – used by System Administrators or system administration tools to perform administration activities. Depending on the task, this user ID must have sa, create object, or primary subscribe permission.
Maintenance User – used by a replicate Replication Server to deliver messages to a replicate data server. This user ID must have the necessary permissions in the replicate data server to execute the commands to which messages to be delivered are mapped.
Replicate user – used by a replicate Replication Server to deliver messages to a primary data server. For delivery for “request” messages, that is, messages from a replicate data server that are selected for delivery to the primary data server, Replication Server uses the user ID of the user who executes the command in the replicate database. This user ID must have the necessary permissions in the primary data server to execute the commands to which messages to be delivered are mapped.
RSI user – used by Replication Server to log in to other Replication Servers to forward messages to be delivered. This user ID must have connect source permission in the replicate Replication Server.
RSSD user – used by Replication Server to log in to the Replication Server System Database (RSSD) that manages its operational data. This user ID must have full control in the RSSD to create and drop objects, execute procedures, and query and update tables.
