Cloudera Edge Management Release Notes
Also available as:
PDF

What's new in Cloudera Edge Management

Cloudera Edge Management is an edge management solution made up of edge agents and an edge management hub. It manages, controls and monitors edge agents to collect data from edge devices and push intelligence back to the edge.

CEM consists of two components:

  • Apache MiNifi is a light-weight edge agent that implements the core features of Apache NiFi, focusing on data collection & processing at the edge.

  • Edge Flow Manager (EFM) is an agent management hub that supports a graphical flow-based programming model to develop, deploy & monitor edge flows on thousands of MiNiFi agents.

CEM provides three main capabilities to the edge flow lifecycle:

Flow Authorship: Edge Flow Manager addresses the challenge of developing IoT applications by offering a code free drag and drop development environment. This development environment offers a NiFi like experience for capturing, filtering, transforming, and transferring data from edge agents to upstream enterprise systems like CDH.

Flow Deployment: Managing the deployment of IoT applications has been an industry challenge. Edge Flow Manager alleviates this challenge by offering a simple, yet powerful, model for deploying applications to agents. Agents registered with Edge Flow Manager are notified when a new or modified application is available. The agents themselve acquire the updated application from Edge Flow Manager to avoid networking complications. The agents will update their local runtime with the application and send verification checks back to Edge Flow Manager once the operation has been verified to have completed successfully.

Flow Monitoring: Agents in CEM send regularly scheduled heartbeats to their Edge Flow Manager instance. The heartbeat represents the most recent snapshot of the agents runtime. Edge Flow Manager stores, analyzes, and renders these heartbeats to end users. The heartbeats allows operators to visualize details such as flow throughput, connection depths, processors running, and overall agent health. Visualizing this information allows for the operator to take appropriate actions where needed.