Pulse provides a scalable architecture for monitoring and analyzing telemetry data from Hadoop-based environments. It supports both Docker- and Kubernetes-based deployments and includes key components such as Agents, Hooks, Connectors, Core Services, Add-ons, and Database Services.
This page explains how to deploy Pulse based on your environment and workload requirements.
Deploy Agents and Hooks
Install Pulse Agents and place Hooks on Hadoop nodes to collect telemetry data and capture query events.
- Deploy Pulse Agents – deploy Agents on Hadoop nodes to gather logs, traces, and metrics.
- Place Hooks – Add Hooks on the nodes to capture service query events.
The approach and steps for installing Agents and placing Hooks remain the same for Docker- and Kubernetes-based deployments.
Deploy Pulse Services
You have two deployment options to deploy the Pulse Services.
Docker-Based Deployment
You can deploy the Pulse Services on dedicated nodes. All Pulse services, including Core services, Add-on services, and databases, run as containers on the Pulse nodes.
Key Characteristics:
- Deploy Pulse Core, Add-on, and Database services on dedicated nodes using Accelo CLI.
- All services (Pulse Core, Add-on, and Databases) are containerized.
For more Information, see Deployment: Docker/Podman-Based.
Kubernetes-Based Deployment
You can deploy the Pulse Services (Core and Add-on) inside a Kubernetes cluster.
Pulse databases are deployed separately on Virtual Machines (VMs) or Bare Metal servers, outside of Kubernetes, to ensure stable and persistent storage.
Key Characteristics:
- Deploy Pulse Core and Add-on services using Pulse Manager Server and Pulse Operator.
- Deploy Pulse Databases on VMs using Pulse Manager Agent.
- Offers better scalability, resilience, and automation via Kubernetes.
For more information, see Deployment: Kubernetes-Based.