@KURAMOTOAHUJATAIRA-3118 , Thank you for the question. Here are some of the pointers that should help understand the difference further:
- Have you verified that the FunctionApp processed all the messages, and the discrepancy is only reported in metric?
- Please check the hosting plan being used and see if one of the limitations is being hit. For example, during scale-out, there's currently a limit of 500 instances per subscription per hour for Linux apps on a Consumption plan. Fo details, see Function App Hosting plan options: Scale.
- The "Diagnose and solve problems" option under the Function App's resource menu contains scenario based troubleshooting step which can be used to get more insights on the Statistics related to function execution. For example, "Availability and performance" has the following scenarios which could help understand more about the execution performances:
- You should also review the "Activity log" under function's resource menu for any issues reported while executing the FunctionApp.
- Also, please review the **Azure Functions reliable event processing **document for best practices related to such reliable event processing.
- The Azure Function Logging could be used to share logs from Function itself using ILogger which can be verified from LA workspace for total execution count. details - Function App Logging
The above are some general guidelines and pointers that should help with the issue. However, if you are specifically facing an issue with your function app that is not showing the correct execution count (even though all the events/inputs were processed), I would suggest creating a support ticket, so that it can be checked from the backend host with other relevant telemetry.
Hope this helps.