Skip to content

Latest commit

 

History

History
102 lines (84 loc) · 5.11 KB

ADOPTERS.md

File metadata and controls

102 lines (84 loc) · 5.11 KB
Raw

Adopters

This is a list of production adopters of OPA (in alphabetical order):

  • bol.com uses OPA for a mix of validating and mutating admission control use cases in their Kubernetes clusters. Use cases include patching image pull secrets, load balancer properties, and tolerations based on contextual information stored on namespaces. OPA is deployed on multiple clusters with ~100 nodes and ~300 namespaces total.

  • Chef integrates OPA to implement IAM-style access control and enumerate user->resource permissions in Chef Automate V2. The integration utilizes OPA's Partial Evaluation feature to reduce evaluation time (in exchange for higher update latency.)

  • Cloudflare uses OPA as a validating admission controller to prevent conflicting Ingresses in their Kubernetes clusters that host a mix of production and test workloads.

  • Intuit uses OPA as a validating and mutating admission controller to implement various security, multi-tenancy, and risk management policies across approximately 50 clusters and 1,000 namespaces. For more information on how Intuit uses OPA see this talk from KubeCon Seattle 2018.

  • Medallia uses OPA to audit AWS resources for compliance violations. The policies search across state from Terraform and AWS APIs to identify security violations and identify high-risk configurations. The policies ingest 1,000s of AWS resources to generate the final report.

  • Netflix uses OPA as a method of enforcing access control in microservices across a variety of languages and frameworks for thousands of instances in their cloud infrastructure. Netflix takes advantage of OPA's ability to bring in contextual information and data from remote resources in order to evaluate policies in a flexible and consistent manner. For a description of how Netflix has architected access control with OPA check out this talk from KubeCon Austin 2017.

  • SAP/InfraBox integrates OPA to implement authorization over HTTP API resources. OPA policies evaluate user and permission data replicated from Postgres to make access control decisions over projects, collaborators, jobs, etc. SAP/Infrabox is used in production within SAP and has several external users.

  • Tremolo Security uses OPA at a London-based financial services company to inject annotations and volume mount parameters into Kubernetes Pods so that workloads can connect to off-cluster CIFS drives and SQL Server instances. Policies are based on external context sourced from OpenUnison. Ability to validate policies offline is a huge win because the clusters are air-gapped. For more information on how Tremolo Security uses OPA see this blog post.

  • Very Good Security (VGS) integrates OPA to implement a fine-grained permission system and enumerate user->resource permissions in their product. The backend is architected as a collection of (polyglot) microservices running on Kubernetes that offload policy decisions to OPA sidecars. VGS has implemented a synchronization protocol on top of the Bundle and Status APIs so that the system can determine when permission updates have propagated. For more details on the VGS use case see this blog post.

In addition, there are several production adopters that prefer to remain anonymous.

  • A Fortune 100 company uses OPA to implement validating admission control and fine-grained authorization policies on ~10 Kubernetes clusters with ~1,000 nodes. They also integrate OPA into their PKI as part of a Certificate RA that serves these clusters.

This is a list of adopters in early stages of production or pre-production (in alphabetical order):

  • Cyral is a venture-funded data security company. Still in stealth mode but using OPA to manage and enforce fine-grained authorization policies.

  • ORY Keto replaced their internal decision engine with OPA. By leveraging OPA, ORY Keto was able to simplify their access control server implementation while retaining the ability to easily add high-level models like ACLs and RBAC. In December 2018, ~850 ORY Keto instances were running in a mix of pre-production and production environments.

Other adopters that have gone into production or various stages of testing include:

If you have adopted OPA and would like to be included in this list, feel free to submit a PR.