Alternative Server
Definition
When defining a spool server, you can define another server as an alternative. This alternative server is used in the following cases:
· The actual server has failed; the system searches for the corresponding alternative server
· The server is overloaded; the system balances the output workload over the actual server and the alternative server. However, to do this, you must activate the Allow Load Balancing option in the definition of the spool server. By default, this option is not activated to ensure that requests are processed in the order in which they are created.
Use
When creating SAP output devices, specify a server under Spool Server in the definition of which an alternative server is specified. In this way you ensure that the requests for this output device are processed even if the primary specified server fails.
You can also distribute the output workload between the spool server and its alternative server by selecting the Allow Load Balancing option in the definition of the spool server.
Server Switch if the Spool Server Fails
If you specify an alternative server when defining a spool server, all output devices are processed by the alternative server if the actual server fails.
The figure below shows a server switch and the specifications required for this in the definition of the spool server:
Distributing the Output Workload if the Spool Server Is Overloaded
The workload of a spool server is calculated from the number of spool work processes, the number of requests to be processes, and the number of pages. If a server is overloaded, the processing of output requests can be delayed. If an alternative server is defined for this server and load balancing is activated, the requests can be distributed between the servers.
The figure below shows load balancing and the specifications required for this in the definition of the spool server:
If you have activated load balancing for a server, the requests to this server may not any longer be processed in the order in which they were created. Requests created later can overtake requests created earlier if they are processed by an alternative server that has a lower workload.
Sequence Consistency has priority in the classification. If an output device is configured to print requests with a consistent sequence; that is, in the order in which they are created, and is assigned to a spool server with load balancing, the load balancing is ignored for this device.
See also:
Transporting the Print Architecture Using Logical Servers
