Engine Routing and Balancing

Engines are dynamically allocated resources. They can migrate among Brokers based on criteria such as load and policy. Use the Engine Balancer to manage logins and re-route Engines to maintain an optimal balance across the grid. The Engine Balancer is a component of the Director. The Primary Director’s balancer always runs. The Secondary Director’s balancer runs only if the Primary is down.

The Director handles routing and load balancing as follows:

1.

The Director regularly polls Brokers for the states of Engines on the Brokers. The Director tests routing mechanisms against each Engine and determines the optimal location for each Engine. Changes in the states of Engines due to load balancing requirements result in changes in the optimal distribution of Engines on Brokers.

2.

The Director sends a request to each Broker that has Engines that must be moved, to log those Engines off.

3.

Engines that must return to their home Broker log off immediately, regardless of the task timeout setting.

4.

Shared Engines that are busy restart immediately without finishing the current task. Engines that are not busy log off immediately.

5.

After an Engine logs off or restarts, it then logs in to the optimal Broker.

Two balancer algorithms are available. Choose one according to how you plan to use the grid:

•

The weight-based balancer algorithm attempts to distribute Engines equally by relative weights, and it also allows rule-based routing using Engine properties.

•

The Home/Shared Balancer routes Engines based on an Engine’s assigned Home Brokers, and the sharing policy of Home Brokers to other Brokers. Both balancers take into account the number of running and pending tasks on each Broker, and the desired maximum and minimum number of Engines for each Broker.

If you change the Engine Balancer on the Director, you must restart it. Also, all balancer settings must be equal on Primary and Secondary Directors. You can configure routing settings for online or offline Brokers.