hadoop-yarn-rest-api 1.1.0

Python wrapper for Hadoop YARN REST API

This is Python Library for Hadoop YARN REST api.

This project follows the [Hadoop YARN REST Api](https://hadoop.apache.org/docs/current/hadoop-yarn/hadoop-yarn-site/ResourceManagerRest.html) document.

We have implemented the following APIs.

• Cluster Information API
• Cluster Metrics API
• Cluster Scheduler API
• Cluster Applications API
• Cluster Application Statistics API
• Cluster Application API
• Cluster Application Attempts API
• Cluster Nodes API
• Cluster Node API

Usage

Install it:

pip install hadoop-yarn-rest-api

Usage:

from hadoop.yarn import Yarn
import xml.etree.ElementTree as ET
import json
import xml.dom.minidom

def main():
    yarn = Yarn("http://localhost:8088", 'json')

    response_obj = yarn.cluster_information()
    response_obj = yarn.cluster_metrics()
    response_obj = yarn.cluster_scheduler()
    response_obj = yarn.cluster_applications()
    response_obj = yarn.cluster_applications({"limit":100})
    response_obj = yarn.cluster_appstatistics()
    response_obj = yarn.cluster_appstatistics({"states":"accepted,running,finished","applicationTypes":"mapreduce"})
    response_obj = yarn.cluster_application("job_id")
    response_obj = yarn.cluster_application_attempts("job_id")
    response_obj = yarn.cluster_nodes()
    response_obj = yarn.cluster_nodes({"states":"RUNNING"})  # NEW, RUNNING, UNHEALTHY, DECOMMISSIONING, DECOMMISSIONED, LOST, REBOOTED, SHUTDOWN
    response_obj = yarn.cluster_node("node_id")

    if yarn.response_type == 'json':
        print(json.dumps(response_obj, indent=4, sort_keys=True))
    elif yarn.response_type == 'xml':
        print(xml.dom.minidom.parseString(ET.tostring(response_obj)).toprettyxml())
    else:
        print(response_obj)

if __name__ == '__main__':
    main()