blob: a2d3b5fdfbeb195df7b18a9b5509a92c97639749 [file]
import logging
import os
import sys
import time
import requests
from apache_ranger.model.ranger_service import RangerService
from apache_ranger.client.ranger_client import RangerClient
# colors only when attached to a terminal (e.g. docker compose with tty), plain text otherwise (e.g. log collectors)
COLOR = sys.stdout.isatty() and 'NO_COLOR' not in os.environ
def style(text, code):
return f'\033[{code}m{text}\033[0m' if COLOR else text
class ConsoleFormatter(logging.Formatter):
LEVELS = {logging.INFO: ('INFO ', '32'), logging.WARNING: ('WARN ', '33'), logging.ERROR: ('ERROR', '1;31')}
def format(self, record):
label, code = self.LEVELS.get(record.levelno, (record.levelname[:5], '0'))
return f"{style(self.formatTime(record, '%H:%M:%S'), '2')} {style(label, code)} {record.getMessage()}"
console_handler = logging.StreamHandler(sys.stdout)
file_handler = logging.FileHandler(os.path.join(os.environ.get('RANGER_ADMIN_LOG_DIR', '/var/log/ranger'), 'create-ranger-services.log'))
console_handler.setFormatter(ConsoleFormatter())
file_handler.setFormatter(logging.Formatter('%(asctime)s %(levelname)-5s %(message)s'))
logging.basicConfig(level=logging.INFO, handlers=(console_handler, file_handler))
RANGER_URL = 'http://localhost:6080'
ranger_client = RangerClient(RANGER_URL, ('admin', os.environ['RANGER_ADMIN_PASSWORD']))
def print_section(title):
print(style(f'━━━ ▶ {title} '.ljust(110, '━'), '1;36'), flush=True)
def wait_for_ranger_admin(timeout=int(os.environ.get('RANGER_ADMIN_WAIT_TIMEOUT', '300'))):
start = time.time()
# readiness is served only to the healthcheck user, authenticated via trusted header (ranger.admin.authn.header.username)
while time.time() - start < timeout:
try:
readiness = requests.get(f'{RANGER_URL}/service/actuator/health/readiness', headers={'X-Forwarded-User': 'healthcheck'}, timeout=5).json()
if readiness.get('status') == 'UP':
return readiness
except (requests.RequestException, ValueError):
pass
logging.info(f'⏳ Waiting for Ranger Admin to become ready: {int(time.time() - start)}s elapsed')
time.sleep(5)
return None
def service_not_exists(service):
# search by name instead of get_service(), which fails with HTTP 404 for a service not created yet
services = ranger_client.find_services({'serviceName': service.name}) or []
return not any(svc.name == service.name for svc in services)
hdfs = RangerService({'name': 'dev_hdfs', 'type': 'hdfs',
'configs': {'username': 'hdfs', 'password': 'hdfs',
'fs.default.name': 'hdfs://ranger-hadoop.rangernw:9000',
'hadoop.security.authentication': 'simple',
'hadoop.security.authorization': 'true',
'policy.download.auth.users': 'hdfs',
'tag.download.auth.users': 'hdfs',
'userstore.download.auth.users': 'hdfs',
'setup.additional.default.policies': 'true',
'default-policy.1.name': 'hive-tez-path',
'default-policy.1.resource.path': '/apps/tez,/tmp/hive',
'default-policy.1.resource.path.is-recursive': 'true',
'default-policy.1.policyItem.1.users': 'hive',
'default-policy.1.policyItem.1.accessTypes': 'read,write,execute',
'default-policy.2.name': 'ranger-audit-path',
'default-policy.2.resource.path': '/ranger/audit',
'default-policy.2.resource.path.is-recursive': 'true',
'default-policy.2.policyItem.1.users': 'rangerauditserver',
'default-policy.2.policyItem.1.accessTypes': 'read,write,execute',
'ranger.plugin.hdfs.policy.refresh.synchronous':'true'}})
hive = RangerService({'name': 'dev_hive', 'type': 'hive',
'configs': {'username': 'hive', 'password': 'hive',
'jdbc.driverClassName': 'org.apache.hive.jdbc.HiveDriver',
'jdbc.url': 'jdbc:hive2://ranger-hive.rangernw:10000',
'hadoop.security.authorization': 'true',
'policy.download.auth.users': 'hive',
'tag.download.auth.users': 'hive',
'userstore.download.auth.users': 'hive',
'ranger.plugin.hive.policy.refresh.synchronous':'true'}})
kafka = RangerService({'name': 'dev_kafka', 'type': 'kafka',
'configs': {'username': 'kafka', 'password': 'kafka',
'zookeeper.connect': 'ranger-zk.rangernw:2181',
'policy.download.auth.users': 'kafka',
'tag.download.auth.users': 'kafka',
'setup.additional.default.policies': 'true',
'default-policy.1.name': 'topic: ATLAS_ENTITIES',
'default-policy.1.resource.topic': 'ATLAS_ENTITIES',
'default-policy.1.policyItem.1.users': 'rangertagsync',
'default-policy.1.policyItem.1.accessTypes': 'create,consume',
'default-policy.2.name': 'consumergroup: ranger_entities_consumer',
'default-policy.2.resource.consumergroup': 'ranger_entities_consumer',
'default-policy.2.policyItem.1.users': 'rangertagsync',
'default-policy.2.policyItem.1.accessTypes': 'consume,describe',
'default-policy.3.name': 'topic: ranger_audit*',
'default-policy.3.resource.topic': 'ranger_audit*,__consumer_offsets',
'default-policy.3.policyItem.1.users': 'rangerauditserver',
'default-policy.3.policyItem.1.accessTypes': 'create,configure,consume,publish,describe,create,delete,alter,describe_configs,alter_configs',
'default-policy.4.name': 'consumergroup: ranger_audit*',
'default-policy.4.resource.consumergroup': 'ranger_audit*',
'default-policy.4.policyItem.1.users': 'rangerauditserver',
'default-policy.4.policyItem.1.accessTypes': 'consume,describe,delete',
'default-policy.5.name': 'cluster: audit server cluster permissions',
'default-policy.5.resource.cluster': '*,dummy',
'default-policy.5.policyItem.1.users': 'rangerauditserver',
'default-policy.5.policyItem.1.accessTypes': 'configure,describe,alter,create,idempotent_write,describe_configs,alter_configs',
'ranger.plugin.audit.filters': "[{'accessResult': 'DENIED', 'isAudited': true},{'resources':{'topic':{'values':['ATLAS_ENTITIES']}},'users':['rangertagsync'],'actions':['create','consume','describe'],'isAudited':false},{'resources':{'consumergroup':{'values':['ranger_entities_consumer']}},'users':['rangertagsync'],'actions':['consume'],'isAudited':false},{'users':['kafka'],'isAudited':false},{'users':['rangerauditserver'],'isAudited':false}]",
'userstore.download.auth.users': 'kafka',
'ranger.plugin.kafka.policy.refresh.synchronous':'true'}})
knox = RangerService({'name': 'dev_knox', 'type': 'knox',
'configs': {'username': 'knox', 'password': 'knox', 'knox.url': 'https://ranger-knox.rangernw:8443',
'policy.download.auth.users': 'knox',
'tag.download.auth.users': 'knox',
'userstore.download.auth.users': 'knox',
'ranger.plugin.knox.policy.refresh.synchronous':'true'}})
yarn = RangerService({'name': 'dev_yarn', 'type': 'yarn',
'configs': {'username': 'yarn', 'password': 'yarn',
'yarn.url': 'http://ranger-hadoop.rangernw:8088',
'policy.download.auth.users': 'yarn',
'tag.download.auth.users': 'yarn',
'userstore.download.auth.users': 'yarn',
'ranger.plugin.yarn.policy.refresh.synchronous':'true'}})
hbase = RangerService({'name': 'dev_hbase', 'type': 'hbase',
'configs': {'username': 'hbase', 'password': 'hbase',
'hadoop.security.authentication': 'simple',
'hbase.security.authentication': 'simple',
'hadoop.security.authorization': 'true',
'hbase.zookeeper.property.clientPort': '2181',
'hbase.zookeeper.quorum': 'ranger-zk.rangernw',
'zookeeper.znode.parent': '/hbase',
'policy.download.auth.users': 'hbase',
'tag.download.auth.users': 'hbase',
'userstore.download.auth.users': 'hbase',
'ranger.plugin.hbase.policy.refresh.synchronous':'true'}})
kms = RangerService({'name': 'dev_kms', 'type': 'kms',
'configs': {'username': 'keyadmin', 'password': os.environ.get('RANGER_KEYADMIN_PASSWORD', 'rangerR0cks!'),
'provider': 'kms://http@ranger-kms.rangernw:9292/kms',
'policy.download.auth.users': 'rangerkms',
'tag.download.auth.users': 'rangerkms',
'userstore.download.auth.users': 'rangerkms',
'ranger.plugin.kms.policy.refresh.synchronous':'true'}})
trino = RangerService({'name': 'dev_trino',
'type': 'trino',
'configs': {
'username': 'trino',
'password': 'trino',
'jdbc.driverClassName': 'io.trino.jdbc.TrinoDriver',
'jdbc.url': 'jdbc:trino://ranger-trino.rangernw:8080',
'policy.download.auth.users': 'trino',
'tag.download.auth.users': 'trino',
'userstore.download.auth.users': 'trino',
'ranger.plugin.trino.policy.refresh.synchronous':'true'
}})
ozone = RangerService({'name': 'dev_ozone',
'type': 'ozone',
'displayName': 'dev_ozone',
'configs': {'username': 'hdfs', 'password': 'hdfs',
'ozone.om.http-address': 'http://om:9874',
'hadoop.security.authentication': 'simple',
'policy.download.auth.users': 'om',
'tag.download.auth.users': 'om',
'userstore.download.auth.users': 'om',
'ranger.plugin.ozone.policy.refresh.synchronous':'true'}})
solr = RangerService({'name': 'dev_solr', 'type': 'solr',
'configs': {'username': 'solr', 'password': 'rangerR0cks!',
'solr.url': 'http://ranger-solr.rangernw:8983',
'setup.additional.default.policies': 'true',
'default-policy.1.name': 'collection: ranger_audits',
'default-policy.1.resource.collection': 'ranger_audits',
'default-policy.1.policyItem.1.users': 'rangerauditserver',
'default-policy.1.policyItem.1.accessTypes': 'query,update',
'policy.download.auth.users': 'solr',
'tag.download.auth.users': 'solr',
'userstore.download.auth.users': 'solr',
'ranger.plugin.super.users': 'solr',
'ranger.plugin.solr.policy.refresh.synchronous':'true'}})
readiness = wait_for_ranger_admin()
print_section('create-ranger-services.py · dev services')
if not readiness:
logging.error('✖ Ranger Admin is not ready, see Ranger Admin log above')
sys.exit(1)
started = os.environ.get('RANGER_ADMIN_START_TIME')
uptime = f' in {int(time.time()) - int(started)}s' if started else ''
service_defs = len(readiness.get('details', {}).get('components', {}).get('service-defs', []))
logging.info(f"✔ Ranger Admin {os.environ.get('RANGER_VERSION', '')} is up and ready{uptime}, {service_defs} service-defs loaded")
services = [hdfs, yarn, hive, hbase, kafka, knox, kms, trino, ozone, solr]
for service in services:
try:
if service_not_exists(service):
ranger_client.create_service(service)
logging.info(f"✔ {service.name} service created")
else:
logging.info(f"{service.name} service already exists")
except Exception as e:
logging.error(f"✖ An exception occurred while creating {service.name} service: {e}")