🪟🔧 AppMonitoringService for custom datadog RUM events

[josephkmh]

What

Adds a monitoring service and hook that allows us to track custom RUM events from the frontend. This is useful for logging when certain code paths are hit, like edge cases or unexpected application states. By logging these events in datadog, we are able to query them and set up monitoring/alerts if thresholds are exceeded.

How

Adds an ApplicationMonitoringService that exposes two methods:

const { trackAction, trackError } = useAppMonitoringService();

trackAction: (actionCode: AppActionCodes, context?: Record<string, unknown>) => void;
Sends a RUM event of type action to datadog

trackError: (error: Error, context?: Record<string, unknown>) => void;
Sends a RUM event of type error to datadog

Both trackAction and trackError can be called with an optional second parameter, context, which is a JSON blob of any other relevant data to log.

In case the datadog RUM SDK is not initialized, trackAction and trackError will log messages in the browser console via console.debug.

Recommended reading order

1. AppMonitoringService.tsx for the implementation
2. LDExperimentService.tsx for an example usage: logging when LaunchDarkly takes too long to respond, and fall back to rendering the app without any Launch Darkly feature/experiment flags.

[timroes]

Code LGTM. Have not tested locally.