Litmus team and CNCF announced today that LitmusChaos has been approved as a CNCF incubating project. For the past two years, it has been a wonderful journey with the community and within CNCF. Growth of usage of Litmus among the enterprises solidifies the fact that the practice of Chaos Engineering is taking its place within DevOps and SRE space. Litmus has grown from <50 installs a day a year ago to 1500+ installs a day by the time it has achieved Incubating status at CNCF, marking a 30x growth is it’s usage.
Litmus installations per day:
Enterprises need a complete Chaos Engineering stack that supports them in their efforts of improving reliability. Though it is convenient to think Chaos Engineering is all about injecting faults, it is much more if you really want to get meaningful results at scale. Apart from fault injection, steady state hypothesis validation and integrating with the existing observability systems are key to achieving results in your Chaos Engineering efforts. Litmus has evolved over a period of three years to cover all the three areas of Chaos Engineering. Litmus 2.0 has the following features:
Fault Injection: Basic chaos experiments are in place at ChaosHub. SREs and DevOps teams use these to construct Chaos Workflows and collaborate on chaos center. ChaosCenter can be deployed for many teams as a single instance to enable sharing of chaos experiments and workflows through the git backend and private ChaosHubs.
Steady state hypothesis validation: Litmus probes are the foundation for building steady state hypothesis validation logic. Litmus probes are flexible, highly granular and declarative. Litmus probes probes feature is one of the strong differentiators for Litmus.
Chaos observability: Litmus provides granular chaos metrics to enable deep analytics around chaos workflows. Application dashboards can be modified easily to add the Litmus chaos context of experiment details and results.
ChaosNative hosts the Litmus control plane for everyone and provides a free forever tier. Litmus community users can simply sign up to the cloud and get a fully featured Litmus control plane instantaneously. Users can then connect their chaos targets to the control plane and run chaos workflows.
To use LitmusChaos Cloud, head over to litmuschaos.cloud and register for a free account.
Once registered, you'll be able to download an Access Key, which will allow you to connect your K8s cluster to the hosted Litmus Control Plane.
Once you have downloaded the Access Key, you will be prompted through the installation step of Litmusctl, the CLI utility tool for managing LitmusChaos resources. Litmusctl will install the Execution Plane components in your K8s cluster in the form of an external agent, which will allow the hosted Control Plane to inject chaos into your cluster.
Once Litmusctl is set up, you can use it to create an external agent in your cluster and connect it to your hosted Control Plane with the steps mentioned in the onboarding page. Finally, once the agent is connected, you are ready to execute your Chaos Workflows.
We can verify the setup by executing the podtato-head demo chaos workflow, which will install the podtato-head application and perform the pod-delete experiment on it. As the workflow is executed, you can visualize the steps of the chaos execution in your dashboard.
Once the execution of the workflow is complete, all the steps should be marked as successful and you should be able to verify the successful execution of the workflow.
You can also check for the ChaosResult and logs for the pod-delete experiment in the Table View.