Protecting the Data Lake
summary
Shows a working integration that puts Open Policy Agent (OPA) in front of Ceph object storage deployed on Kubernetes with Rook, where the RADOS gateway queries OPA and OPA returns JSON decisions (not just allow/deny) to enforce attribute policies such as location-based bucket access (Bob in USA can read an USA bucket; Alice in UK cannot). Also outlines Kafka and Elasticsearch adapters, the Rego language and an interactive play.openpolicyagent.org tutorial; claims meet strict latency and availability requirements.
AI-written summary · about
chapters (14)
- 0:00Start
- 1:16Data is King!
- 2:50Data Lake Features • Controlled Content
- 5:03Data Lake Platform: Kafka
- 6:38Data Lake Platform: Ceph
- 7:34Data Lake Platform: Elasticsearch
- 8:27Security Challenge Overview
- 11:09OPA: Community
- 12:01OPA: General-purpose policy engine
- 13:07OPA: Features • Decorative Policy Language Regol
- 14:40How does OPA work?
- 15:473 Steps to OPA
- 23:14Ceph Data Protection: Setup
- 26:26Demo: Ceph Data Protection
related talks
34mFine-grained Authorization in a Containerized World
25mPolicing Your Kubernetes Clusters with Open Policy Agent (OPA) - Mark Puddick & Amith Nambiar
19mDynamic Authorization and Policy Control for Your Kubernetes Cluster
33mOpen Policy Agent Deep Dive
29mIntroduction to Open Policy Agent
45mOpen Policy Agent Deep Dive
38mPolicy as Code with Open Policy Agent
25mPolicing Your Kubernetes Clusters with Open Policy Agent (OPA)
