Configure Pulse to Monitor Flink

This page describes how to configure Pulse to monitor Flink performance and enable application- and job-level observability.

Before You Begin

Before deploying the ad-flink connector to process Flink data, make sure your Hadoop cluster is already configured to send Flink metrics to Pulse. For detailed steps, see Configure ODP Flink.

Deploy the Flink Connector Add-on

Deploy the ad-flink-connector add-on service to poll job-related data from the YARN Resource Manager and Flink Job History Server.

The connector retrieves data such as job start time, execution statistics, and vertex details, which Pulse uses to display detailed job information.

Docker-Based Pulse Deployment

If you have deployed Pulse with Docker, follow the steps below to deploy the add-on service.

1. On a Pulse node, run the following command.

accelo deploy addons

The list of available Pulse add-on services appears on screen.

2. Select the Flink Connector along with other components or services you want to deploy.

? Select the components you would like to install:  

  [x] Flink Connector

  [ ] Actions Service  

  [x] Alerts (Agents MUST be configured)  

  [ ] CONFIG SERVER DB  

  [x] Core Connectors  

  [x] Dashplot  

  [ ] Director (Agents MUST be configured)  

​

  [Use arrows to move, space to select, <right> to all, <left> to none, type to filter]

Kubernetes-Based Pulse Deployment

If you have deployed Pulse on Kubernetes, see Deploy Pulse Add-on Services to deploy the Flink Connector add-on service.