#!/usr/bin/env python3 import argparse import subprocess import sys from datetime import datetime, timedelta from typing import List import requests from checker import nagios def get_disk_wwn_ids(ignore_non_smart: bool = False) -> List[str] or bool: wwn_ids = [] try: output = subprocess.check_output(["sudo", "smartctl", "--scan"]) for line in output.decode("utf-8").strip().split("\n"): parts = line.split() if len(parts) >= 3: device = parts[0] device_type = parts[2].replace('scsi', 'sat,auto') try: smart_supported = subprocess.check_output(["sudo", "smartctl", "-i", device, "-d", device_type]).decode("utf-8") if "SMART support is: Enabled" in smart_supported: wwn_line = [line for line in smart_supported.split("\n") if "LU WWN Device Id" in line] wwn_id = '0x' + wwn_line[0].replace('LU WWN Device Id: ', '').replace(' ', '') wwn_ids.append(wwn_id) # else: # # TODO: warn if a drive doesn't support SMART except subprocess.CalledProcessError as e: if ignore_non_smart: continue else: print(f"UNKNOWN: subprocess Error - {e}") sys.exit(nagios.UNKNOWN) except subprocess.CalledProcessError as e: print(f"UNKNOWN: subprocess Error - {e}") sys.exit(nagios.UNKNOWN) return wwn_ids def get_smart_health(wwn_id: str, scrutiny_endpoint: str) -> dict: url = f"{scrutiny_endpoint}/api/device/{wwn_id}/details" response = requests.get(url) if response.status_code == 200: return response.json() elif response.status_code == 404: print(f"Disk {wwn_id} not found on Scrutiny") return {} else: print(f"Scrutiny Error {response.status_code} for disk {wwn_id}: {response.text}") return {} def main(args): results = {} wwn_ids = get_disk_wwn_ids(args.ignore_non_smart) metics_out_of_date = False for wwn_id in wwn_ids: smart_health = get_smart_health(wwn_id, args.scrutiny_endpoint) name = f'/dev/{smart_health["data"]["device"]["device_name"]} {wwn_id}' # differentiate disks in RAID arrays results[name] = { 'wwn_id': wwn_id, 'failed_attributes': [], } metadata = smart_health['metadata'] # For testing # smart_health['data']['device']['UpdatedAt'] = '2023-04-28T23:00:03.071184465Z' last_updated = datetime.strptime(smart_health['data']['device']['UpdatedAt'][:-4] + 'Z', '%Y-%m-%dT%H:%M:%S.%fZ') if datetime.utcnow() - timedelta(hours=args.time_delta_limit) > last_updated: metics_out_of_date = True if smart_health and len(smart_health['data']['smart_results']): try: disk_data = smart_health['data']['smart_results'][0]['attrs'] except Exception as e: print('UNKNOWN - failed to parse data:', e) print('Key "data":', smart_health['data'].keys()) print('Key "smart_results":', len(smart_health['data']['smart_results'])) print('Key "attrs":', smart_health['data']['smart_results'][0].keys()) sys.exit(nagios.UNKNOWN) for attribute_id, values in disk_data.items(): if values['status'] == 0: continue # elif values['status'] == 2 and not args.warn_non_critical: # continue values['attribute_name'] = metadata[attribute_id]['display_name'] values['metadata'] = metadata[attribute_id] if 'observed_thresholds' in values['metadata'].keys(): del values['metadata']['observed_thresholds'] results[name]['failed_attributes'].append(values) else: results[name]['status'] = 'no data' crit_disks = {} warn_disks = {} for disk, values in results.items(): if values['status'] != 'no data': for item in values['failed_attributes']: if item['status'] == 2 and args.warn_non_critical: if disk not in warn_disks.keys(): warn_disks[disk] = [] warn_disks[disk].append({ 'raw_value': item['raw_value'], 'display_name': item['metadata']['display_name'] }) if item['status'] == 4: if disk not in crit_disks.keys(): crit_disks[disk] = [] crit_disks[disk].append({ 'raw_value': item['raw_value'], 'display_name': item['metadata']['display_name'] }) else: if disk not in warn_disks.keys(): warn_disks[disk] = [] warn_disks[disk].append({ 'raw_value': 'no data', 'display_name': 'no data' }) dt = '