liveBook · Manning

Karpenter

As explained in Effective Platform Engineering

Karpenter is an open-source Kubernetes node provisioning tool that dynamically provisions and consolidates compute capacity.

Karpenter dynamically provisions and consolidates Kubernetes compute capacity, allowing user workloads to run beyond a fixed management node group. section 7.4.6

Use Kubernetes scheduling constraints to control the capacity it supplies: node affinity and selectors pull workloads toward matching nodes, while tolerations admit workloads onto tainted, dedicated nodes. section 7.4.6

How Effective Platform Engineering covers it

How deep it goes
Demonstrates scheduling constraints and dynamic node-pool tests but does not detail provisioning or consolidation internals.
Examples
Tests AMD64 and ARM workloads, management-node placement, and Bats assertions for Running pods.
What it assumes
Assumes Kubernetes and EKS familiarity while demonstrating node selectors, affinity, taints, and tolerations through manifests.
What you can do after
Diagnose dynamic capacity by deploying architecture-specific workloads and checking whether each reaches Running status.
How it connects
Connects Kubernetes scheduling with node affinity, node selectors, tolerations, taints, and node pools.
Cover of Effective Platform Engineering
liveBookmanning.com
Read Effective Platform Engineering on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Effective Platform Engineering. The section links point to the parts of the book it draws on.