Pulse Add-on Services

Pulse Add-on services are optional services to extend observability for specific Hadoop services.

You can choose which add-ons to deploy based on the Hadoop services in your environment and your monitoring needs.

These services are categorized by purpose, such as:

This classification helps you understand the function of each service and deploy only those that are relevant to your setup.

Service	Description
Acceldata SQL Analysis Service	Provides Hive table-level analysis. The data is available after the fsimage is successfully loaded.
Core Connector	This service comprises two components: ad-connectors and ad-sparkstats.\n\n\n- ad-connectors: Processes and analyzes all Hive, Tez, and MapReduce queries, as well as YARN applications.\n- ad-sparkstats: Processes asd analyzes Spark data.
Impala Connector	Processes and analyzes Impala service data.
Oozie Connector	Processes and analyzes Oozie service data.
Ozone Connector	Processes and analyzes Ozone service data.
Druid Analyzer	Processes and analyzes Druid service data. Runs on port 19090.
Kafka Connector (Both)	Processes and analyzes Kafka service data.
Trino Connector	Processes and analyzes Trino service data.
ClickHouse Connector	Processes and analyzes ClickHouse service data.
Flink Connector	Processes and analyzes Flink service data.
Airflow Connector	Processes and analyzes Airflow service data.
Standalone Connector	Deploys `ad-connector` as a standalone service.

Service	Description
FSAnalytics V2	`ad-fsanalyticsv2-connector`: Fetches and processes the FSImage, then pushes the data to ElasticSearch.
FS Elastic	Runs FSAnalytics on a separate node and routes statistics or logs into a dedicated ElasticSearch instance.
GraphQL HA	Enables high availability (HA) for the Pulse UI.
LDAP	Integrates with the organization’s LDAP for user authentication and RBAC in Pulse UI.
Yarn Optimizer	Used for optimizing the YARN Cluster for resources.
Alfred	Used to process and analyze fingerprints of YARN applications.
Recommendation Service	Used for calculating recommendations for various Hadoop services for query recommendations.
Pulse Profiler (ad-pulse-profiler)	Used for polling, processing, and analyzing JVM profiling data.
Anomaly Detection	Used to process time-series data and display anomalous data points.
Old UI	Enables you to access the Pulse legacy UI.