Chaos Engineering for VMware environments

A good amount of the enterprise IT infrastructure runs on VMware. Increase in digital traffic requires the business services to remain highly available to the end users. The best operating practise in such cases is to continuously validate the service reliability and optimise the recovery procedures. Faults of various types are designed and injected into VMware infrastructure and application recovery is validated. Service Level Objectives are continuously monitored by the ops teams while chaos is in play at virtual infrastructure. ChaosNative Litmus offers various chaos experiments for VMware infrastructure using which chaos workflows are designed and executed.

Use Cases

  • Continuous validation of SLOs when VMware vSphere is the underlying infrastructure

  • Evaluate the resilience of Kubernetes infrastructure that is built on VMware vSphere

  • Continuous chaos testing of virtual infrastructure in CI pipelines

  • Validate the functioning of monitoring and observability platforms built on virtual machine infrastructure

  • DR validation scenarios of business services built with virtual machine infrastructure

vmware

Some of the chaos experiments available for VMWare environment:

Chaos Experiment

Description

Tunable Parameters

Disk Loss Experiment

Description

VMWare Disk Loss experiment causes the detachment of Virtual Disks attached to a VM for a specified duration of time and later attaches them back to the VM.

Tunables

Necessary Inputs:

  • Disk IDs

  • Disk VM MOIDs



Secrets Inputs:

  • vCenter Server URL

  • vCenter Username

  • vCenter Password

VM PowerOff Experiment

Description

VMware VM Power-off experiment causes power-off of the VMWare VM(s) provided by the Moid(s) of the instance.

Tunables

Necessary Inputs:

  • VM MOIDs



Secrets Inputs:

  • vCenter Server URL

  • vCenter Username

  • vCenter Password

Process Kill Experiment

Description

VMWare Process Kill experiment kills target processes in a Linux-based VMware VM

Tunables

Necessary Inputs:

  • Process IDs

  • VM Name



Secrets Inputs:

  • vCenter Server URL

  • vCenter Username

  • vCenter Password

  • VM Username

  • VM Password

Service Kill Experiment

Description

VMware Service Kill experiment kills given services running on a Linux VM for a specified chaos interval before restarting it. It also validates the self-healing behaviour of the services, if specified.

Tunables

Necessary Inputs:

  • Service Names

  • VM Name


Tunable Inputs:

  • Self healing Service (enable /disable)


Secrets Inputs:

  • vCenter Server URL

  • vCenter Username

  • vCenter Password

  • VM Username

  • VM Password

Host Reboot Experiment

Description

VMWare Host Reboot experiment causes a target VMWare ESX Host to reboot. It can also validate the behaviour of hosts belonging to HA clusters.

Tunables

Necessary Inputs:

  • Host Name

  • Host Datacenter


Tunable Inputs:

  • High Availability Cluster (enable /disable)


Secrets Inputs:

  • vCenter Server URL

  • vCenter Username

  • vCenter Password

HTTP Chaos Experiment

Description

VMware HTTP Chaos experiment will induce HTTP chaos into the VMWare VM(s) for the specified duration. It uses toxiproxy to inject HTTP chaos.

Tunables

Necessary Inputs:

  • Toxicity

  • Latency

  • Request Timeout


Tunable Inputs:

  • Stream Port

  • Stream Type

  • Listen Port

  • Stream URL

  • Listen URL

  • HTTP Chaos Type

  • Operating System

  • Rate Limit

  • Data Limit


Secrets Inputs:

  • vCenter Server URL

  • vCenter Username

  • vCenter Password

  • VM Username

  • VM Password

Get a demo of the chaos scenarios for VMWare

Get a demo
Get a demo of the chaos scenarios for VMWare