Optimizing CI/CD Pipeline with Shift-Left Enterprise Observability

Containers and microservices have increased the speed at which software is put into production. They have also increased scalability, especially for Cloud-native applications. Agile development and CI/CD practices have been instrumental in making these new application modernization methods effective. For production applications, i.e. once the application code and enhancements have been released through the CD process, Enterprise Observability (EO) and Machine Learning (ML) are extremely effective in detecting and isolating application issues for SRE, DevOps and Ops team in production. Enterprise Observability is the combination of Metrics, Events, Traces and Logs into a rich mosaic of rich contextual data. EO also provides precise information at 1 second intervals, with no sampling. Machine Learning learns from that rich contextual data to expose potential issues and enable alerts for those issues. Many users have asked us if EO/ML could be beneficial early in the CI/CD pipeline. The answer is YES! In this presentation we’ll identify: How and where EO/ML can be applied to the CI/CD pipeline Who would benefit from the CI/CD pipeline enhancements (hint: Developers and DevOps, QA) Overview the research being applied to EO/ML for the CI/CD pipeline such as fault injection Shift-Left EO/MLwill provide many benefits for software teams, such as: Hardening software for production release more rigorously than Unit and Automated Test can do Help detect “unknown unknowns” in the CI/CD pipeline Define and provide recovery guidelines for problems identified using Machine Learning Enable higher confidence that your software will be more robust when you go live.