Skip to content

Dynatrace Observability Lab: Predictive Auto-Scaling for Kubernetes Workloads#

Support Policy

This is a demo project created by the Developer Relations team at Dynatrace, showcasing integrations with open source technologies.

Support is provided via GitHub issues only. The materials provided in this repository are offered "as-is" without any warranties, express or implied. Use them at your own risk.

View the Code

The code for this repository is hosted on GitHub. Click the "View Code on GitHub" link above.

Struggling to keep up with the demands of dynamic Kubernetes environments? Manual scaling is not only time-consuming and reactive but also prone to errors. In this demo we harness the power of Dynatrace Automations and Davis AI to predict resource bottlenecks and automatically open pull requests to scale applications. This proactive approach minimizes downtime, helps you to optimize resource utilization, and ensures your applications perform at their best.

We achieve this by combining predictive AI to forecast resource limitations with generative AI to modify Kubernetes manifests on GitHub by creating pull requests for scaling adjustments. If you'd like a closer look at how this works, you can run the full demo on your own tenant.

Considerations and Limitations#

While this demo showcases the power of automating Kubernetes scaling, it's important to be aware of a few aspects to ensure smooth integration into your specific environment:

  • GitOps Deployment Assumptions: For the automatic scaling pull requests to function effectively, the demo makes some assumptions about your GitOps setup:
    • The Kubernetes Auto Remediation workflows will only target workloads with these specific annotations. This behavior could be changed but was introduced to speed up workflow runs and save resources.
      • predictive-kubernetes-scaling.observability-labs.dynatrace.com/enabled: 'true'
  • GitHub Codespace Usage: If you follow the above instructions, a GitHub Codespace will be created under your account. While running this demo in a GitHub Codespace is free for most users due to generous usage limits and default billing settings, we recommend deleting the Codespace after completing the tutorial to avoid potential future charges if you exceed the free tier. To delete your Codespace, go to https://github.com/codespaces. For more information, see the GitHub Codespaces documentation.
  • Davis CoPilot API Usage: This demo utilizes Davis CoPilot API calls for its generative AI capabilities. These calls might cause costs in the future.

Compatibility#

Deployment Tutorial Compatible
Dynatrace Managed ❌
Dynatrace SaaS ✔️