Intricacies involved in Kubernetes application deployment and monitoring setup across various cloud hosting platforms create a higher entry barrier. The need for an enterprise-ready container orchestration application environment, proven to be scalable, reliable and resilient has brought LitmusChaos and Kublr together. Developers can now seize the fear of chaos and move towards a resilient future.

What’s on the table for you?

The ease of cluster deployment accompanied by integrated centralized monitoring and logging in Kublr blended with the principle of “Chaos first” held by LitmusChaos, a Chaos engineering framework creates an indefectible amalgamation. Enabling large-scale cloud-native applications to have a production-grade environment setup. Chaos testing must be performed iteratively with the incremental application development to enhance its reliability. Software development life cycle (SDLC) for cloud applications should not isolate the development from chaos testing. LitmusChaos enables developers and SREs to work in an intra-mutual ecosystem rather than a conventional strategy of isolated chaos testing performed at the end of application development.

The disaster recovery strategy endorses the use of single-plane setup for the swift response and rebound policy. Kublr is a one-stop solution for container management. Empowering developers with customized application delivery and management workflows. For Kublr users, LitmusChaos provides chaos injection feature for analysing the behaviour of the production-grade application in the event of chaos in the development phase itself. Thereby enhancing the application architecture and making it immune to any breakdown. For LitmusChaos users Kublr provides a simplified container setup with hassle-free monitoring and logging integration. Such bilateral conjunction caters to the requirements of scalable enterprise cloud solution.

LitmusChaos on Kublr

For setting up a sock-application to demonstrate the working infrastructure of Kublr platform and LitmusChaos the steps are given below:

• Setup the Kublr platform with central monitoring and logging enabled on any cloud hosting platform as shown in figure 1 & 2.
• Download the kubeconfig file from the Kublr cluster and attach it with your CLI
• Clone the litmuschaos/litmus repository
• Go to litmus/monitoring/platforms/kublr
• Now you can follow the readme for setting up the LitmusChaos on Kublr, annotating the litmus and sock-shop metrics so that they can be scrapped by the Kublr’s Prometheus as shown in figure 3.
• Add the Grafana dashboard and execute the chaos experiments as shown in figure 4 & 5.
• You can also set up Argo workflow (optional)

Figure 1: Kublr’s full platform cluster

Figure 2: Kublr cluster stats

Figure 3: Kublr’s Prometheus

Figure 4: Node and Pod Chaos metrics dashboard in Kublr’s Grafana

Figure 5: Sock Shop Performance dashboard in Kublr’s Grafana

Conclusion and Resources

Augmentation of any community is based upon its inclusive expansionism. Kublr-LitmusChaos integration pays the way for a future where secure cloud applications are devices with a minimal time frame. Watch Kubernetes Chaos Engineering with MayaData and Kublr webinar to learn more. For a taste of chaos engineering, do try litmus demo and start your chaos engineering journey today. The contributing members of LitmusChaos have been engaging and collaborating with other Kubernetes focused products for endowing a sound cloud-native layout. Here is another pillar added to aid the developers in using LitmusChaos on Okteto easing out the barrier of system requirements and limitations for development environments on the cloud.

Join the Community

Are you an SRE or a Kubernetes enthusiast? Does Chaos Engineering excite you?

Join our community on Slack for detailed discussion, feedback & regular updates On Chaos Engineering For Kubernetes,
To join, our slack please follow the following steps!

1. Join the Kubernetes slack
2. Join the #litmus channel on the Kubernetes slack

Check out the LitmusChaos GitHub repository and do share your feedback:
Submit a pull request if you identify any necessary changes.