In this episode, we welcome back Max Henchel, Technical Consultant at Microsoft, to dive deep into the topic of telemetry. We previously discussed Dual-write, but realized we needed to go back to basics to explain the telemetry infrastructure and how to actually start measuring performance in a smart way.
Max is one of the primary drivers behind the ISM Telemetry Model, a framework that simplifies working with Application Insights for Dynamics users.
In this episode, we discuss:
- Why isn't out-of-the-box telemetry enough? How the ISM model acts as a "wrapper" to make it easier for partners and customers to get started without drowning in data.
- The importance of custom emission points: Why you can't just "flip a switch," but actually need to design your telemetry to understand critical business processes like order flows and batch jobs.
- Cost vs. Benefit: Strategies for data storage and how to keep costs down in Azure by focusing on short-term analysis (e.g., 30-day retention).
- From "It's slow" to hard facts: How to use telemetry to address subjective experiences of performance with actual data and system "heartbeats".
- Exception logging: How to find the notorious "Object reference not set to an instance of an object" error using call stacks directly in App Insights, without needing to copy production databases for debugging.
- Cross-Enterprise Monitoring: Using a single Application Insights repository to track an end-to-end chain across different applications like AIS or Power Platform.
Three tips to get started:
- Set it up now: It takes less than an hour to get basic telemetry in place.
- Use FastTrack reports: Leverage ready-made dashboards to get immediate visualization.
- Learn Kusto (KQL): Max recommends the Kusto Detective Agency for a fun, interactive way to learn how to query your data.
