Copyright © TIBCO Software Inc. All Rights Reserved
Copyright © TIBCO Software Inc. All Rights Reserved


Chapter 6 Fault Tolerance and Load Balancing : Fault Tolerance and Load Balancing for the Interior Server

Fault Tolerance and Load Balancing for the Interior Server
The Interior Server can be deployed to provide both for fault tolerance and load balancing. Machines that belong to one group provide for fault tolerance within that group, while machines in different groups provide for load balancing among these groups.
Figure 16 Interior Server: Load Balancing and Fault Tolerance
Figure 16 shows a configuration with three fault tolerant groups, where servers are grouped as follows:
Fault Tolerance for the Interior Server
Fault tolerance is achieved since each server in a group acts as a backup for the other server in the same group. Multiple groups can coexist, each containing two or more servers. In each of these groups, servers are started in a sequence so that the server that starts first works until it fails.
Upon the first server’s failure, the second server installed in the same group takes its role, and so on. Servers have no primary or secondary functions, and the order in which they provide fault tolerance is based on the order in which they were started.
Load Balancing and Public Smart Routing for the Interior Server
Load balancing for the Interior component is achieved by adding multiple Interior engines or multiple fault tolerant groups.
Using the Public Smart Routing feature, you can distribute the workloads and alleviate the likelihood of bottlenecks by using multiple server cluster based on preset rules (predefined set of criteria). A rule-based routing mechanism makes decisions, based on a combination of configurable conditions and dispatches workloads to the best fitting cluster for processing.
To learn more, see Public Smart Routing.
Configuring the Interior Server
The Interior Server must be configured with the following information:
Checkpoint Database   This database contains all checkpoints (transaction execution records) from an Interior component. This way, in the case of the machine’s failure, these saved transaction records are transferred from the disabled machine to the one that takes over its function.
When you initially set up TIBCO BusinessConnect, the connection for the Checkpoint Database is set and is named bc-check-point-db by default.
You can modify the configuration of the Checkpoint database and any other databases by clicking the connection alias.
All required parameters   The Interior component parameters, such as Service, Network, Daemon Host, Daemon Port, Heartbeat Interval, Activation Interval, and Activation Delay, must be set to support fault tolerance.
These parameters are set during deployment.
For more details, see TIBCO BusinessConnect Interior Server Administration, "Configuring Interior Server.par."

Copyright © TIBCO Software Inc. All Rights Reserved
Copyright © TIBCO Software Inc. All Rights Reserved