Spotfire Server and Environment - Quick Start

Routing introduction

The Spotfire Server provides routing capabilities within the Spotfire environment.

The Spotfire Server routes clients (end users, scheduled update jobs and Automation Services jobs) to the appropriate service instance, based on the default routing or configured routing rules.

Note: The routing capability of Spotfire Server features built-in load balancing, enabling analyses to be routed based on resource utilization, session affinity and custom resource pool routing rules, to the proper Web Player service instances. Therefore, you should not add any external load balancer between Spotfire Server and the nodes.

The routing algorithm: focus on resource utilization

By default, when a user tries to open an analysis, or a scheduled update for an analysis is triggered, Spotfire will try to route the request to the instance with the best health status (available/strained/exhausted), where the analysis is already loaded, and which has the lowest number of user sessions.

That is, that the analysis is loaded is used as an affinity condition. The reason for this is that much of the data in memory can be reused, so it is usually more effective to route many users using the same analysis into the same instance, rather than redirecting them to different instances. When analysis data is shared between users, additional users on the same instance accessing the analysis file will have a low impact on performance. For more information, see Understanding routing.

After an analysis is opened in a client, all subsequent requests by the same user for the session are forwarded to the instance that was used for the initialization; this way, the Spotfire Server routing maintains analysis session affinity.

Resource pools and routing rules

In addition to the default routing algorithm, administrators can create custom routing rules and resource pools:

  • A resource pool is a logical group of Spotfire Web Player instances. Each Web Player instance can only be part of one resource pool.
  • A routing rule specifies the resource pool on which an analysis opens.

When you have specified a custom routing rule, the affected users are routed to the target resource pool (the specified Web Player instances) as a first step. After that, the default routing algorithm is applied.

Resource pools and routing rules are used together to control the quality of service for different users, groups or analyses. Routing rules can be specified for users, groups, or specific analysis files, and are defined and applied in priority order, similar to mail sorting rules. Routing rules can be re-prioritized, enabled, disabled, and re-mapped to a different resource pool.

See Routing rules and Resource pools for more information.

Default routing rule

If there are no resource pools configured or there is no applicable routing rule, the default routing rule is applied.

On initial installation, the default routing rule includes all the Web Player instances that are available in your Spotfire implementation, but you can modify this default routing rule if desired. See The default routing rule for more information.

Also, administrators can add scheduled update rules to routing rules that apply to analysis files. That way, the administrator can have the analysis pre-loaded on selected instances in a resource pool, and have the analysis refreshed at specified intervals.

See Routing rules and Creating resource pools for more information.