How to: Monitor Apps in Azure App Service
App Service provides built in monitoring functionality in the Azure Portal. This includes the ability to review quotas and metrics for an app as well as the App Service plan, setting up alerts and even scaling automatically based on these metrics.
Although this article refers to web apps, it also applies to API apps and mobile apps.
Understanding Quotas and Metrics
Applications hosted in App Service are subject to certain limits on the resources they can use. The limits are defined by the App Service plan associated with the app.
If the application is hosted in a Free or Shared plan, then the limits on the resources the app can use are defined by Quotas.
If the application is hosted in a Basic, Standard or Premium plan, then the limits on the resources they can use are set by the size (Small, Medium, Large) and instance count (1, 2, 3, ...) of the App Service plan.
Quotas for Free or Shared apps are:
- Amount of CPU allowed for this application in a 3-minute interval. This quota re-sets every 3 minutes.
- Total amount of CPU allowed for this application in a day. This quota re-sets every 24 hours at midnight UTC.
- Total amount of memory allowed for this application.
- Total amount of outgoing bandwidth allowed for this application in a day. This quota re-sets every 24 hours at midnight UTC.
- Total amount of storage allowed.
The only quota applicable to apps hosted on Basic, Standard and Premium plans is Filesystem.
More information about the specific quotas, limits and features available to the different App Service SKUs can be found here: Azure Subscription Service Limits
If an application in its usage exceeds the CPU (short), CPU (Day), or bandwidth quota then the application will be stopped until the quota re-sets. During this time, all incoming requests will result in an HTTP 403.
If the application memory quota is exceeded, then the application will be re-started.
If the Filesystem quota is exceeded, then any write operation will fail, this includes writing to logs.
Quotas can be increased or removed from your app by upgrading your App Service plan.
Metrics provide information about the app, or App Service plan's behavior.
For an Application, the available metrics are:
- Average Response Time
- The average time taken for the app to serve requests in ms.
- Average memory working set
- The average amount of memory in MiBs used by the app.
- CPU Time
- The amount of CPU in seconds consumed by the app. For more information about this metric see: CPU time vs CPU percentage
- Data In
- The amount of incoming bandwidth consumed by the app in MiBs.
- Data Out
- The amount of outgoing bandwidth consumed by the app in MiBs.
- Http 2xx
- Count of requests resulting in a http status code >= 200 but < 300.
- Http 3xx
- Count of requests resulting in a http status code >= 300 but < 400.
- Http 401
- Count of requests resulting in HTTP 401 status code.
- Http 403
- Count of requests resulting in HTTP 403 status code.
- Http 404
- Count of requests resulting in HTTP 404 status code.
- Http 406
- Count of requests resulting in HTTP 406 status code.
- Http 4xx
- Count of requests resulting in a http status code >= 400 but < 500.
- Http Server Errors
- Count of requests resulting in a http status code >= 500 but < 600.
- Memory working set
- Current amount of memory used by the app in MiBs.
- Total number of requests regardless of their resulting HTTP status code.
For an App Service plan, the available metrics are:
App Service plan metrics are only available for plans in Basic, Standard and Premium SKU.
- CPU Percentage
- The average CPU used across all instances of the plan.
- Memory Percentage
- The average memory used across all instances of the plan.
- Data In
- The average incoming bandwidth used across all instances of the plan.
- Data Out
- The average outgoing bandwidth used across all instances of the plan.
- Disk Queue Length
- The average number of both read and write requests that were queued on storage. A high disk queue length is an indication of an application that might be slowing down due to excessive disk I/O.
- Http Queue Length
- The average number of HTTP requests that had to sit on the queue before being fulfilled. A high or increasing HTTP Queue length is a symptom of a plan under heavy load.
CPU time vs CPU percentage
There are 2 metrics that reflect CPU usage. CPU time and CPU percentage
CPU Time is useful for apps hosted in Free or Shared plans since one of their quotas is defined in CPU minutes used by the app.
CPU percentage on the other hand is useful for apps hosted in basic, standard and premium plans since they can be scaled out and this metric is a good indication of the overall usage across all instances.
Metrics Granularity and Retention Policy
Metrics for an application and app service plan are logged and aggregated by the service with the following granularities and retention policies:
- Minute granularity metrics are retained for 48 hours
- Hour granularity metrics are retained for 30 days
- Day granularity metrics are retained for 90 days
Monitoring Quotas and Metrics in the Azure Portal.
You can review the status of the different quotas and metrics affecting an application in the Azure Portal.
Quotas can be found under Settings>Quotas. The UX allows you to review: (1) the quotas name, (2) its reset interval, (3) its current limit and (4) current value.
Metrics can be access directly from the resource blade. You can also customize the chart by: (1) click on it, and select (2) edit chart. From here you can change the (3) time range, (4) chart type, and (5) metrics to display.
You can learn more about metrics here: Monitor service metrics.
Alerts and Autoscale
Metrics for an App or App Service plan can be hooked up to alerts, to learn more about this, see Receive alert notifications
App Service apps hosted in basic, standard or premium App Service Plans support autoscale. This allows you to configure rules that monitor the App Service plan metrics and can increase or decrease the instance count providing additional resources as needed, or saving money when the application is over-provision. You can learn more about auto scale here: How to Scale and here Best practices for Azure Monitor autoscaling
If you want to get started with Azure App Service before signing up for an Azure account, go to Try App Service, where you can immediately create a short-lived starter web app in App Service. No credit cards required; no commitments.
- For a guide to the change from Websites to App Service see: Azure App Service and Its Impact on Existing Azure Services