Skip to main content Link Menu Expand (external link) Document Search Copy Copied
Azure Chaos Studio Labs

!!! This Lab is NOT complete yet as of 4/19/23, it’s being worked on as time permits, stay tuned !!!

Overview

Faults fall into two categories; service-direct or agent-based. Agent-based faults requires the installation of the Chaos Agent, unlike a service-direct fault. A CPU Pressure fault is an agent-based fault, so we will need to install the Chaos Agent Extenstion on the VM.

For more details on the Chaos Agent click here}

Step 1 - Create a Resource Group

  1. Sign in to Azure portal

  2. Select resource groups, then click create
    For the Resouce Group name, I used rg-CPU-Chaos-Lab3, use whatever name you prefer.

  3. Select subscription and enter values and finish creating the Resource Group and use rg-NSG-Chaos-Lab1 for the Resource Group name.

Step 2 - Create the VM

  1. While in the Azure Portal navigate to Home > Create a resouce then Create a VM

  2. Use the the resource group you created in Step 1.

  3. For the VM name enter “vm-cpu-exp” and keep everything else as simple as possible for the VM to save costs. Here are my suggestions, but you can use whatever you like: 
      Availability options = No infrastructure reduancy required
  Security type = Standard
  Image = Windows 10 Pro
  Size = Standard_B1ls - 1 vcpu, 0.5 GiB memory
  OS disk type = Standard SSD (locally-redundant storage)
  4. Finsh creating the VM.

Step 2 - Create a Managed Identity

  1. While in the Azure Portal, enter Managed Identities into the Search Bar then click on Manage Identities

  2. Now, click on Create and populate the properties with the following details:

      Subscription = <Select your Subscription>
  Resource Group = <Select Resource Group created in Step 1>
  Region = <Select your Region, I am using East US>
  Name = vm-chaos-manaeddid
  Size = Standard_B1ls - 1 vcpu, 0.5 GiB memory

    Now, finish up by clicking on Review + create

Important Note - The managed identity is required as it allows the experiement to communicate with the Chaos Agent on the VM.

Step 3 - Enable the Target & Capabilities

Now that we have the VM and Managed Identity created we can enable the target and capabilities for the VM.

  1. Open Chaos Studio and and click on Targets

Step 4 - Create the Experiement

Step 5 - Run the Experiement

Step 6 - Monitor the CPU Pressure

Note: Technically, Main.bicep does not need to be passed any parameters, by default it will set values for the experimentPrefix and networkSeucruityGroupPrefix if you do not pass them.

  1. As the Main.bicep is being deployed, you notice the status displays as /Running. When it has completed, it will render some JSON data to the screen. Make note of the property provisioningState if it was successful it will have a value of Succeded.

  2. Now, navigate to the Experiments in Chaos Studio and search for an experiement with a prefix of exp-bicep-lab2 and click Start to run the experiment. Did the experiement run or did it fail?

Take a few minutes to try and figure out how to fix the issue and what you might change in the Bicep to fix it. Hint: Marksman often do what to maintain their accuracy?