Key concepts of DataPower Interact Gateway analytics

An overview of how DataPower Interact Gateway analytics works.

How DataPower Nano Gateway and Analytics component work together

DataPower Nano Gateway generates event records for MCP server and LLM API requests that are processed by the IDIG platform. The event records contain information about each request, such as request metadata, application identifiers, response status, and processing details. DataPower Nano Gateway temporarily buffers event records and sends them in batches to the Analytics component for storage and processing. Analytics stores the event data and makes it available for reporting and analysis.

By default, an IDIG cluster deploys Analytics component in the base configuration. For more information about the Analytics component profile, see Deployment profiles. For more information about how to enable and disable the Analytics component, see Deploying IDIG cluster.

Accessing your analytics data

In DataPower Interact Gateway UI, click Analytics to view the analytics data. The maximum number of events that can be accessed from the UI is 10,000. The maximum number of event records that can be displayed per page in the UI is 200. For more details, see Accessing analytics in the DataPower Interact Gateway UI.

Note: If the analytics dashboard does not show MCP server and LLM API events that you expect, the most common cause is a communication failure between your DataPower Nano Gateway and analytics component. Check both the DataPower Nano Gateway operator and IDIG cluster logs.

Offloading your analytics data to a third-party system

The event data that is received from your DataPower Nano Gateway can be directed to a third-party system, instead of storing in the analytics component. For more information, see Offloading analytics data.

How the analytics data is stored for your MCP server and LLM APIs

In IBM API Studio, the Telemetry policy controls the level of detail that is recorded in the event data. The log options available in the Telemetry policy are as follows. For details on how to configure policies in IBM API Studio, see Configuring policies.

Telemetry verbosity levels are as follows:

none - No attributes are included except verbosity.
basic - Includes attributes that are provided in the configurations and policy definitions along with the required attributes per OpenTelemetry specification.
normal - Includes request and response headers in addition to attributes included for basic level.
detailed - Includes request_body and response_body attributes containing the truncated payload of request and response respectively in addition to what's included for normal level.