Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the challenges of building a `kubectl blame` command in this conference talk from KubeCon + CloudNativeCon Europe 2021. Delve into the complexities of Kubernetes control loops and their impact on tracing state changes. Learn from the Tilt team's attempts to create a tool for assigning blame across cluster updates, and gain insights into labels, informers, UIDs, owner refs, events, and the internal workings of kubectl apply. Discover the limitations of various approaches, including naive implementations, kubectl rollout, helm upgrade, kubespy trace, and Tilt up. Examine the fundamental aspects of declarative systems and cascading Kubernetes specs while understanding what Kubernetes provides for developers seeking to interpret API objects for human comprehension.
Syllabus
Intro
Outline
What this talk is NOT
Control Loops
Declarative Systems are
The Sample App
Naive Approach (crash)
kubectl rollout (crash)
helm upgrade --wait (crash)
helm upgrade --wait --debug
kubespy trace (crash)
Tilt up (crash)
Rant Time
What Does Kubernetes Give You?
Cascading Kubernetes Specs
Taught by
CNCF [Cloud Native Computing Foundation]