Monitor an Azure installation

This article is applicable to customers with infrastructure deployed in Microsoft Azure that requires monitoring - for example, a host running one or more web apps. In these cases, the goal is to quickly detect and alert on any issues users have reaching the host or accessing its applications. In addition, monitoring results help answer the questions:

  • What sort of app performance are my users experiencing? (Experience)
  • What paths are user-to-app traffic taking? (Delivery)
  • Are my service providers living up to their SLAs? (Delivery)
  • How do I determine where network problems are originating? (Delivery)

What does this guide cover

  • The recommended approach to monitoring an Azure installation with AppNeta Performance Manager (APM).
  • Where and how to deploy Monitoring Points.
  • How to configure the APM components (Experience and Delivery) to monitor an Azure installation.
  • How to get notified when a network or application issue occurs.
  • How to gain insight into how your installation is performing.

AppNeta recommends deploying a Monitoring Point within the same Azure Virtual Network (VNet) as your host as well as to each location that interacts with the host. This might include Headquarters, Branch offices, and Home offices. If customers access your web apps, you can use AppNeta Global Monitoring Points (GMPs) as a proxies for customers in various regions. Once the Monitoring Points are deployed, they can be configured to emulate users accessing the hosted applications (Experience) and to monitor network health (Delivery).

Network diagram showing Monitoring Points in Headquarters, a Branch office, and a home office with paths to a host and an AppNeta CMP located in Azure.

In this example, when Experience monitoring is configured, there is a web path and a single-ended network path created to the host from remote Monitoring Points in the cloud, Headquarters, and the Branch office. Separate dual-ended network paths are created via Delivery monitoring from Monitoring Points in each remote location to the Container-based Monitoring Point (CMP) deployed in Azure to provide bi-directional network performance monitoring. These paths provide the following:

  • Web paths - enable you to determine that a web app can be accessed and how it is performing.
  • Single-ended paths - enable you to determine that the host can be accessed as well as the routing and performance of the network path to the host.
  • Dual-ended paths - enable you to determine that the VNet can be accessed as well as the routing and performance of the network paths both to and from the VNet.

Note that the Native Monitoring Point (NMP), deployed in the Home office, does not support Experience monitoring so there is no path from it to the host.

Step 1: Deploy Monitoring Points

  1. Deploy a CMP to the VNet where your web apps are hosted.
  2. If customers access your web apps, set up a Global Monitoring Point in each region you have customers.
  3. Deploy Monitoring Points in the Headquarterss, Branch Offices, and Home offices you want to monitor application and network performance from. The Monitoring Points should be connected to the same subnet/segment as users in these locations in order to monitor from a user perspective.

The following are typical Monitoring Point deployments:

Step 2: Emulate web app users

By emulating a user, Experience monitoring helps you answer the questions:

  • Is the web app running?
  • What sort of app performance are my users experiencing?

It allows you to:

  • Monitor key web apps to identify any issues affecting end user experience.
  • Compare user complaints about poor web app performance against a consistent baseline.
  • Use associated single-ended network paths to see the path taken to access your web apps and to determine whether the host they run on can be accessed.

Experience monitoring prerequisites

Monitoring Points should be deployed with interface(s) connected to the applicable (typically end-user) subnets.

Experience monitoring procedure

  1. Create a Web App Group for each web app you want to monitor.
    1. Specify each Monitoring Point and interface you want to monitor as test sources.
    2. Use the web app URL as the test target.
    3. Add a Browser workflow (using Selenium scripting) that exercises the web app. At a minimum, it should simply access the web app. Web paths are created for all combinations of source, target, and workflow.
  2. Tag web paths in order to group them in ways that make sense to your business.

Note that for each web path created, an associated single-ended network path is also created.

Step 3: Monitor network health

The combination of web paths and their associated single-ended network paths enable you to determine whether web app performance problems are due to network issues or application issues. What’s missing is network monitoring on the return path. For return path monitoring, dual-ended paths are created to the CMP deployed in Azure alongside the host.

By creating dual-ended paths, Delivery monitoring can help you answer the questions:

  • Can I access the VNet that my host is deployed on?
  • How is the network between my user and my web app performing (in both directions)?
  • Are my service providers living up to their SLAs?
  • How do I determine where network problems are originating?

Delivery monitoring prerequisites

Monitoring Points should be deployed with interface(s) connected to the applicable (typically end-user) subnets.

Delivery monitoring procedure

  1. Create a Path Template Group to monitor the health of the network between the remote Monitoring Points and the VNet your host is using.
    1. Use the IP address of the CMP deployed in the VNet as the target.
    2. Specify “Dual Ended” paths.
    3. Add source interfaces from each remote Monitoring Point (typically the “Auto” interface) to create paths.
  2. Tag network paths in order to group them in ways that make sense to your business.

Step 4: Set up alert notifications

Consider who will need to be notified in real time when issues are detected, what systems they use to manage alerts, and how to integrate AppNeta notifications with those systems. AppNeta Performance Manager supports notification via:

  • Email - Use this method if you don’t have any other event monitoring infrastructure or if you prefer email alerts. Set up using the Update Notification Options page in APM.
  • Event integration - Use this method if you already have an event monitoring system in place. Integrate directly with that system via POSTs that contain JSON event payloads.
  • SNMP notifications - Use this method if you are integrating with an SNMP system. Set up using the Manage SNMP page in APM.

Adjusting alert profiles if required

Default alert profiles are assigned to web paths and network paths when they are created. If you are receiving too many or too few alerts you can create or modify alert profiles as necessary.

Step 5: Analyze monitoring results

In general, if you notice performance issues on web paths, you should also review the associated network paths. If the network paths look okay then the problem is at the application, otherwise there is a network issue that needs to be investigated further.

When you receive alert notifications, use the following procedures to investigate the cause:

Use dashboards for an “at a glance” way to view the status of your web app and your network.

  • Application Quality dashboard - enables you to view the performance of the network and web paths that deliver an application over time (up to 30 days).
  • Web App Violation Summary dashboard - provides an at-a-glance view of the most significant web path violations over a selected period.
  • Network Violation Summary dashboard - provides an at-a-glance view of the most significant network path violations over a selected period.
  • Current Network Violation Map dashboard - provides a geographical view of your current network status. With this dashboard, issues occurring on multiple paths in a particular geographic region are easily identified.

Schedule reports to be delivered via email to those in your organization interested in web app and/or network performance.

  • Application Quality report - enables you to view the performance of the network and web paths that deliver an application over time (up to 1 year). It combines raw data and violations sourced from Experience and Delivery across a selection of applications, networks, and geographical locations to enable a high-level overview of the performance and trends over time. This report can scale up to span quarterly results and includes a one page summary aimed at an executive audience.
  • Location Bandwidth Quality report - compares the performance of a WAN network path to the stated performance of the internet service package you purchased from your ISP. It enables you to determine whether your service providers are living up to their SLAs.