Ranger-5081: CI: Add check to verify plugin installation in ranger-service containers

[ChinmayHegde24]

Currently, with GitHub CI, the pipeline brings up all the containers that are supported by dev-support/ranger-docker.

What changes were proposed in this pull request?

Adding a way to verify that the ranger plugin setup succeeded when the containers come up as another check in the CI pipeline.

[kumaab]

I recommend the use of Python language for adding checks that are more readable, maintainable, reusable and extendable. It's also preferable for extensive JSON parsing.

Please take a look at below sample code (AI generated) as a starting point, can be further improvised:

#!/usr/bin/env python3

"""
Check Ranger plugin status via Ranger Admin API.
This script is intended to be used as a CI check step, and follows
a function-based, maintainable structure for easy extension.
"""

import os
import sys
import time
import json
from typing import List
import requests
from dotenv import load_dotenv

def load_env(env_path: str):
    """Load environment variables from a .env file."""
    load_dotenv(dotenv_path=env_path)

def trigger_knox_activity(knox_user: str, knox_pass: str, knox_endpoint: str):
    """Trigger activity for KNOX to make plugin active."""
    print("\nTriggering Knox activity to ensure plugin status is updated...")
    try:
        requests.get(knox_endpoint, auth=(knox_user, knox_pass), verify=False, timeout=10)
        print("Knox activity triggered.")
    except Exception as e:
        print(f"Warning: Knox trigger failed: {e}")

def fetch_plugin_info(ranger_admin_user: str, ranger_admin_pass: str, endpoint: str):
    """Fetch plugin info from Ranger Admin API."""
    print(f"\nFetching plugin info from {endpoint} ...")
    try:
        resp = requests.get(endpoint, auth=(ranger_admin_user, ranger_admin_pass), timeout=10)
        resp.raise_for_status()
        return resp.json()
    except Exception as e:
        print(f"Failed to fetch plugin info: {e}")
        return None

def check_plugin_status(response: list, expected_services: List[str]) -> bool:
    """Check the status of plugins for expected services."""
    print("\n<---------  Plugin Status  ---------->")
    failed = False
    for svc in expected_services:
        print(f"\nChecking service type: {svc}")
        entries = [entry for entry in response if entry.get("serviceType") == svc]
        count = len(entries)
        if count == 0:
            print(f"MISSING: No plugins found for service type '{svc}'.")
            failed = True
            continue
        active_plugins = [
            entry for entry in entries
            if entry.get("info", {}).get("policyActiveVersion")
        ]
        active_count = len(active_plugins)
        print(f"\U0001F7E2 Active plugins: {active_count} / {count} total plugins found.")
        if active_count == 0:
            print(f"WARNING: Plugins present but NONE are active for '{svc}'.")
            failed = True
        print("Details:")
        for entry in entries:
            host = entry.get("hostName", "unknown")
            app_type = entry.get("appType", "unknown")
            active_ver = entry.get("info", {}).get("policyActiveVersion", "null")
            print(f"- Host: {host}, AppType: {app_type}, PolicyActiveVersion: {active_ver}")
    return not failed

def main():
    # Load .env from the parent directory
    script_dir = os.path.dirname(os.path.abspath(__file__))
    env_path = os.path.join(script_dir, '..', '.env')
    load_env(env_path)

    RANGER_HOST = os.getenv("RANGER_HOST", "http://localhost:6080")
    ENDPOINT = f"{RANGER_HOST}/service/public/v2/api/plugins/info"

    KNOX_USER = os.getenv("KNOX_USER")
    KNOX_PASS = os.getenv("KNOX_PASS")
    KNOX_ENDPOINT = os.getenv("KNOX_ENDPOINT", "https://localhost:8443/gateway/sandbox/webhdfs/v1/?op=LISTSTATUS")

    RANGER_ADMIN_USER = os.getenv("RANGER_ADMIN_USER")
    RANGER_ADMIN_PASS = os.getenv("RANGER_ADMIN_PASS")

    expected_services = ["hdfs", "hbase", "kms", "yarn", "kafka", "ozone", "knox", "hive"]

    # 1. Trigger knox activity
    trigger_knox_activity(KNOX_USER, KNOX_PASS, KNOX_ENDPOINT)

    # 2. Wait for status update
    time.sleep(60)

    # 3. Fetch plugin info
    response = fetch_plugin_info(RANGER_ADMIN_USER, RANGER_ADMIN_PASS, ENDPOINT)
    if not response or not isinstance(response, list):
        print("No plugin info returned from API.")
        sys.exit(1)

    # 4. Check status
    all_ok = check_plugin_status(response, expected_services)
    print()
    if not all_ok:
        print("\u274C One or more plugins are missing or inactive.")
        sys.exit(1)
    else:
        print("\u2705 All expected plugins are present and active.")

if __name__ == "__main__":
    main()