This article answers frequently asked questions (FAQ) about common issues that can occur when you use Application Insights availability monitoring in Application Insights.
My site looks OK, but I experience test failures. Why is Application Insights alerting me?
Starting from December 2023, Availability Test web requests use TLS 1.3 in supported regions to align with the latest security standards. If your tests fail, check whether your web server accepts inbound TLS 1.3 client connections. For regional support details, see Application Insights Availability Tests TLS 1.3 Enablement.
Does your test have the Parse dependent requests option enabled? That forces a strict check on resources such as scripts and images. These kinds of failures might not be noticeable in a browser. Check all images, scripts, style sheets, and any other files that are loaded by the page. If any of them fails, the test is reported as failed, even if the main HTML page loads without issues. To filter the test for such resource failures, clear the Parse dependent requests check box in the test configuration.
To reduce the odds of noise from transient network blips and other distractions, make sure that Enable retries for test failures configuration is selected. You can also test from other locations and manage the alert rule threshold for those locations to prevent location-specific issues that cause undue alerts.
To see details about why the failure was reported, select any of the red dots from the availability scatter plot experience, or any availability failure from the search results. The test results, together with the correlating server-side telemetry (if enabled), should help you understand why the test failed. Common causes of transient issues are the network status and connectivity.
Did the test time out? We abort tests after two minutes. If your ping or multi-step test takes longer than two minutes, we report it as a failure. Consider breaking the test into multiple tests that can finish more quickly.
Did all locations report failure, or only some of them? If only some locations reported failures, this situation might occur because of network or content delivery network (CDN) issues. Again, select any of the red dots from the availability scatter plot experience to better understand why the location reported failures.
Why didn't I get an email message when the alert triggered or resolved (or both)?
Check the action group configuration for the alert to verify that your email account is directly listed, or that a distribution list you're on is configured to receive notifications. If the distribution list is set up for notifications, then check the distribution list configuration to verify that it can receive external email. Also, check to see whether your mail administrator has policies configured that might cause this issue to occur.
Why didn't I receive the webhook notification?
Make sure that the application that receives the webhook notification is available, and that it successfully processes the webhook requests. For more information, see Webhook actions for log alert rules.
I'm getting "403 - Forbidden" errors. What does this mean?
This error indicates that you have to add firewall exceptions to allow the availability agents to test your target URL. For a full list of agent IP addresses to allow, see the IP exception article.
Why do I see "Other values" in my web test telemetry?
Application Insights limits availability tests to 100 per Application Insights resource. If you create custom web test results by calling the TrackAvailability() API, you could attempt to exceed that limit. For any availability test name, location, or other field values where the cardinality exceeds 100, its value is replaced with "Other values". If you need more than 100 different web test names or locations, you can use a second Application Insights component.
Why don't I see any related server-side telemetry to diagnose test failures?
This situation might exist if you have Application Insights set up for your server-side application and sampling is in operation. Select a different availability result.
Can I call code from my web test?
No. The steps of the test must be in the .webtest file. Also, you can't call other web tests or use loops.
Is there a difference between "web tests" and "availability tests"?
The two terms can be used interchangeably. "Availability tests" is a more generic term that includes the single URL ping tests in addition to the multi-step web tests.
My multi-step test doesn't finish. Why?
This failure might occur if the test run reaches the size limit of 300 KB. The test can also fail if it uses unsupported features, such as loops, data sources, or references to other web tests.
My multi-step test doesn't upload. Why?
Multi-step web tests are limited to 100 requests per test. Also, the test is stopped if it runs longer than two minutes.
Can I run a test by using client certificates?
No. This scenario isn't currently supported.
Why does my availability test continue to run after it is paused or disabled?
It is an expected behavior. Depending on how close the deadline for the scheduled execution of the next iteration is, availability tests may not be stopped in time before they perform the next iteration. In the same way, when you create, disable, or manipulate new tests, there is a delay.
How can I allow other users to view my availability test results?
To enable other users to view your availability test results in Application Insights, you must grant them at least read access to the resource group containing your Application Insights resources. Availability tests are separate Azure resources, identified as Microsoft.Insights/webtests. They are created within the same resource group as your Application Insights resources. For detailed steps on how to control access, see Control access in the resource group.
Contact us for help
If you have questions or need help, create a support request, or ask Azure community support. You can also submit product feedback to Azure feedback community.