Search code examples
azureazure-functionsperformance-testingazure-application-insightsazure-api-management

Azure application insights availability from multiple test locations - Are requests fired sequentially or all at one?


I have set-up a standard Availability test to my Azure Functions and APIM in Azure, using the standard availability feature in Application Insights. Azure advises to use a minimum of 5 test locations, i.e. the places from where the Azure servers send web requests to my URL:

Our minimum number of recommended test locations is five to ensure that you can distinguish problems in your website from network issues

Source: https://learn.microsoft.com/en-us/azure/azure-monitor/app/availability-standard-tests

My questions: Are the tests coming from these 5 locations all fired at once (concurrent processing) or one-by-one (sequentially)? Since the processing of indivual requests by my API could take up to 30 seconds, I want to know how to interpret the average processing times stated in the test results. Concurrent processing can take much longer than sequential processing and taking the average.

Thanks!


Solution

  • (I am a part of Application Insights team)

    Answering the question - neither sequentially nor all at once.

    Every location runs tests completely independently from other locations. On top of it, the first test execution is randomized within the first 5 minutes (and after it a test is run every 5 minutes).