Mirror Servers (SQL Anywhere)

Sybase SQL Anywhere (v12 and higher) supports database mirroring through the use of mirror servers. PowerDesigner models mirror servers as extended objects with a stereotype of <<MirrorServer>>.

Creating a Mirror Server

You can create a mirror server in any of the following ways:

Select Model > Mirror Servers to access the List of Mirror Servers, and click the Add a Row tool.
Right-click the model (or a package) in the Browser, and select New > Mirror Server.

Mirror Server Properties

You can modify an object's properties from its property sheet. To open a mirror server property sheet, double-click its diagram symbol or its Browser entry in the Mirror Servers folder.

The following extended attributes are available on the Options tab:

Name	Description
Type	Specifies the type of mirror server to create. You can choose between: Primary - defines a virtual or logical server, whose name is the alternate server name for the database, which can be used by applications to connect to the server currently acting as the primary server. There can be only one PRIMARY server for a database. Mirror - defines a virtual or logical server, whose name is the alternate server name for the database, which can be used by applications to connect to the server currently acting as the read-only mirror. There can be only one MIRROR server for a database. Arbiter - assists in determining which of the PARTNER servers takes ownership of the database. The arbiter server must be defined with a connection string that can be used by the partner servers to connect to the arbiter. There can be only one ARBITER server for a database. Partner - is eligible to become the primary server and take ownership of the database. You must define two PARTNER servers for database mirroring, and both must have a connection string and a state file. In a read-only scale-out system, you must define one PARTNER server. This server is the root server, and runs the only copy of the database that allows both read and write operations. Copy - In a read-only scale-out system, this value specifies that the database server is a copy node. All connections to the database on this server are read-only. You do not have to explicitly define copy nodes for the scale-out system; you can choose to have the root node define the copy nodes when they connect. Scripting name: Type
Using auto parent	[copy only] Specifies that the primary server will assign a parent for this server. Scripting name: UsingAutoParent
Parent	[copy only] Specifies a tree of servers for a mirroring or scale-out system and indicates the servers from which the non-participating nodes obtain transaction log pages. Scripting name: ParentServer
Alternate parent	[copy only] Specifies an alternate parent for the copy node. Scripting name: AlternateParentServer
Primary	[copy only] Specifies that the parent server is the primary server. Scripting name: PrimaryParentServer
Connection string	Specifies the connection string to be used to connect to the server. Scripting name: ConnectionString
Log file	Specifies the location of the log file that is sent between mirror servers. Scripting name: LogFile
Preferred	[partner only] Specifies whether the server is the preferred server in the mirroring system, which assumes the role of primary server whenever possible. Scripting name: Preferred
State file	[arbiter, partner] Specifies the location of the file used for maintaining state information about the mirroring system. Scripting name: StateFile