Open Policy Agent Deep Dive
summary
Open Policy Agent (OPA) is a host‑local, in‑memory policy engine that returns arbitrary JSON decisions (not just booleans) and integrates with systems like Kubernetes (admission control), SSH, Docker and Terraform. Policies are evaluated via HTTP POST with JSON input/output to named policies, external context is provided by data injected via HTTP PUT into OPA, and the language supports iteration via implicit variables, about 50 built‑in operators, and a distinction between functions vs virtual documents for reusable logic and indexed lookups. Performance tools include tracing, indexing and unit tests; built‑ins are side‑effect free.
AI-written summary · about
chapters (14)
- 0:00Start
- 2:46Use OPA to policy-enable your project
- 4:05How Policies are invoked
- 6:53Example: HTTP API Authorization
- 10:59Lookup and Compare Values
- 12:35Create Rules
- 14:47Create Functions
- 27:07Policies Use Context
- 31:11What about Arrays?
- 39:37Iterate over Everything
- 40:50Create a Virtual Document
- 41:31Different Syntaxes for Virtual Sets
- 42:16Create Virtual Dictionaries too
- 43:15Virtual Docs support iteration Functions dont Dictionary
related talks
33mOpen Policy Agent Deep Dive
41mOpen Policy Agent (OPA) Intro & Deep Dive
45mOPA Deep Dive
29mOpen Policy Agent (OPA) Intro & Deep Dive
17mOpen Policy Agent Intro
29mIntroduction to Open Policy Agent
38mPolicy as Code with Open Policy Agent
19mDynamic Authorization and Policy Control for Your Kubernetes Cluster