The PrivOps Matrix Architecture
A Next Generation platform
The PrivOps Matrix is a control plane & data catalog you can deploy anywhere: AWS, other cloud providers and on premise. Our modular approach to data governance, automation integration and real-time analytics makes it possible to securely and compliantly create new applications and services an order of magnitude faster than with traditional methods.
The control plane of the PrivOps Matrix consists of a set of standard functions and libraries that handle workflow automation, time series versioning of everything in the metadata catalog , data ETL and integration. Integration is greatly simplified since developers need not have any knowledge of how the Matrix stores its data or the underlying database. The Matrix uses node.js, the most popular and quickest growing open source server-side programming language. As a result, integrations have access to nearly 500,000 plug-ins making it easy to connect to virtually any API, software, database, or tool in existence.
The metadata catalog contains the resource, policy, function, and data objects within the PrivOps Matrix. Each object is stored as time series data for audit and troubleshooting; organizations can demonstrate every change, every event, every access at every point in time. The catalog is built on top of distributed open source Apache Cassandra. Apple, for example, runs more than 78,000 Cassandra nodes as the backbone for their technology platform. At IBM, The Weather Company processes over 30 billion API requests daily on its Cassandra based IoT platform.
PrivOps Matrix nodes can be located anywhere – cloud providers and/or on-premise and because Cassandra provides for replication across datacenters, disaster recovery is built in by design.
Constructing Applications with the PrivOps Matrix
Example: GDPR personal data filtering/masking application
Each object in the catalog is a time series metadata (or actual data) object. Using automation workflows, objects can be updated and/or linked to any other set of objects in real time, making it possible to reconfigure applications in real time.
Because of the modular nature of the PrivOps Matrix approach to application deployment, it’s a simple matter to deploy applications across multiple clouds. It’s so simple, it can be automated based on an event trigger, like a change in pricing by a cloud provider. Only the data components or objects of an application stack specific to a cloud (or on-premise) are switched out as part of the deployment flow.
Since all objects are stored as time series, versioning and audit are automatic. In addition, since all the application components are modular and isolated, change management is simplified because components specific to one location are isolated from components specific to other locations.