def execute(args: typing.NamedTuple):
aztk = Aztk()
cluster_id = args.cluster_id
if not args.force:
confirmation_cluster_id = input(
"Please confirm the id of the cluster you wish to delete: ")
if confirmation_cluster_id != cluster_id:
log.error(
"Confirmation cluster id does not match. Please try again.")
return
if aztk.client.delete_cluster(cluster_id):
log.info("Deleting cluster %s", cluster_id)
else:
log.error("Cluster with id '%s' doesn't exist or was already deleted.",
cluster_id)
def execute(args: typing.NamedTuple):
aztk = Aztk()
log.info('-------------------------------------------')
log.info('spark cluster id: {}'.format(args.cluster_id))
log.info('username: {}'.format(args.username))
log.info('-------------------------------------------')
if args.ssh_key:
ssh_key = args.ssh_key
else:
ssh_key = aztk.client.secrets_config.ssh_pub_key
ssh_key, password = utils.get_ssh_key_or_prompt(ssh_key, args.username,
args.password,
aztk.client.secrets_config)
aztk.client.create_user(cluster_id=args.cluster_id,
username=args.username,
password=password,
ssh_key=ssh_key)
if password:
log.info('password: %s', '*' * len(password))
elif ssh_key:
log.info('ssh public key: %s', ssh_key)
log.info('-------------------------------------------')
def execute(args: typing.NamedTuple):
aztk = Aztk()
# read cluster.yaml configuartion file, overwrite values with args
cluster_conf = ClusterConfig()
cluster_conf.merge(
uid=args.cluster_id,
size=args.size,
size_low_pri=args.size_low_pri,
vm_size=args.vm_size,
wait=args.wait,
username=args.username,
password=args.password,
docker_repo=args.docker_repo)
print_cluster_conf(cluster_conf)
if cluster_conf.custom_scripts:
custom_scripts = []
for custom_script in cluster_conf.custom_scripts:
custom_scripts.append(
aztk_sdk.spark.models.CustomScript(
script=custom_script['script'],
run_on=custom_script['runOn']
)
)
else:
custom_scripts = None
jars_src = aztk_sdk.utils.constants.DEFAULT_SPARK_JARS_SOURCE
# create spark cluster
cluster = aztk.client.create_cluster(
aztk_sdk.spark.models.ClusterConfiguration(
cluster_id=cluster_conf.uid,
vm_count=cluster_conf.size,
vm_low_pri_count=cluster_conf.size_low_pri,
vm_size=cluster_conf.vm_size,
custom_scripts=custom_scripts,
docker_repo=cluster_conf.docker_repo,
spark_configuration=aztk_sdk.spark.models.SparkConfiguration(
spark_defaults_conf=os.path.join(
aztk_sdk.utils.constants.DEFAULT_SPARK_CONF_SOURCE, 'spark-defaults.conf'),
spark_env_sh=os.path.join(
aztk_sdk.utils.constants.DEFAULT_SPARK_CONF_SOURCE, 'spark-env.sh'),
core_site_xml=os.path.join(
aztk_sdk.utils.constants.DEFAULT_SPARK_CONF_SOURCE, 'core-site.xml'),
jars=[
os.path.join(jars_src, path) for path in os.listdir(jars_src)
]
)
),
wait=cluster_conf.wait
)
if cluster_conf.username:
ssh_key = aztk.client.secrets_config.ssh_pub_key
ssh_key, password = utils.get_ssh_key_or_prompt(
ssh_key, args.username, args.password, aztk.client.secrets_config)
aztk.client.create_user(
cluster_id=cluster_conf.uid,
username=cluster_conf.username,
password=password,
ssh_key=ssh_key
)
if cluster_conf.wait:
log.info("Cluster %s created successfully.", cluster.id)
else:
log.info("Cluster %s is being provisioned.", cluster.id)
def print_cluster_conf(cluster_conf):
log.info("-------------------------------------------")
log.info("spark cluster id: %s", cluster_conf.uid)
log.info("spark cluster size: %s",
cluster_conf.size + cluster_conf.size_low_pri)
log.info("> dedicated: %s", cluster_conf.size)
log.info("> low priority: %s", cluster_conf.size_low_pri)
log.info("spark cluster vm size: %s", cluster_conf.vm_size)
log.info("custom scripts: %s", cluster_conf.custom_scripts)
log.info("docker repo name: %s", cluster_conf.docker_repo)
log.info("wait for cluster: %s", cluster_conf.wait)
log.info("username: %s", cluster_conf.username)
if cluster_conf.password:
log.info("Password: %s", '*' * len(cluster_conf.password))
log.info("-------------------------------------------")
def execute(args: typing.NamedTuple):
aztk = Aztk()
ssh_conf = SshConfig()
ssh_conf.merge(cluster_id=args.cluster_id,
username=args.username,
job_ui_port=args.jobui,
job_history_ui_port=args.jobhistoryui,
web_ui_port=args.webui,
jupyter_port=args.jupyter,
name_node_ui_port=args.namenodeui,
host=args.host,
connect=args.connect)
http_prefix = 'http://localhost:'
log.info("-------------------------------------------")
log.info("spark cluster id: %s", ssh_conf.cluster_id)
log.info("open webui: %s%s", http_prefix, ssh_conf.web_ui_port)
log.info("open jobui: %s%s", http_prefix, ssh_conf.job_ui_port)
log.info("open jobhistoryui: %s%s", http_prefix,
ssh_conf.job_history_ui_port)
log.info("open jupyter: %s%s", http_prefix, ssh_conf.jupyter_port)
log.info("open jupyter: %s%s", http_prefix,
ssh_conf.name_node_ui_port)
log.info("ssh username: %s", ssh_conf.username)
log.info("connect: %s", ssh_conf.connect)
log.info("-------------------------------------------")
# get ssh command
try:
ssh_cmd = utils.ssh_in_master(
client=aztk.client,
cluster_id=ssh_conf.cluster_id,
webui=ssh_conf.web_ui_port,
jobui=ssh_conf.job_ui_port,
jobhistoryui=ssh_conf.job_history_ui_port,
namenodeui=ssh_conf.name_node_ui_port,
jupyter=ssh_conf.jupyter_port,
username=ssh_conf.username,
host=ssh_conf.host,
connect=ssh_conf.connect)
if not ssh_conf.connect:
log.info("")
log.info(
"Use the following command to connect to your spark head node:"
)
log.info("\t%s", ssh_cmd)
except batch_error.BatchErrorException as e:
if e.error.code == "PoolNotFound":
raise aztk_sdk.error.AztkError(
"The cluster you are trying to connect to does not exist.")
else:
raise
def execute(args: typing.NamedTuple):
aztk = Aztk()
jars = []
py_files = []
files = []
if args.jars is not None:
jars = args.jars.replace(' ', '').split(',')
if args.py_files is not None:
py_files = args.py_files.replace(' ', '').split(',')
if args.files is not None:
files = args.files.replace(' ', '').split(',')
log.info("-------------------------------------------")
log.info("Spark cluster id: %s", args.cluster_id)
log.info("Spark app name: %s", args.name)
log.info("Wait for app completion: %s", args.wait)
if args.main_class is not None:
log.info("Entry point class: %s", args.main_class)
if jars:
log.info("JARS: %s", jars)
if py_files:
log.info("PY_Files: %s", py_files)
if files:
log.info("Files: %s", files)
if args.driver_java_options is not None:
log.info("Driver java options: %s", args.driver_java_options)
if args.driver_library_path is not None:
log.info("Driver library path: %s", args.driver_library_path)
if args.driver_class_path is not None:
log.info("Driver class path: %s", args.driver_class_path)
if args.driver_memory is not None:
log.info("Driver memory: %s", args.driver_memory)
if args.executor_memory is not None:
log.info("Executor memory: %s", args.executor_memory)
if args.driver_cores is not None:
log.info("Driver cores: %s", args.driver_cores)
if args.executor_cores is not None:
log.info("Executor cores: %s", args.executor_cores)
log.info("Application: %s", args.app)
log.info("Application arguments: %s", args.app_args)
log.info("-------------------------------------------")
aztk.client.submit(cluster_id=args.cluster_id,
application=aztk_sdk.spark.models.AppModel(
name=args.name,
application=args.app,
application_args=args.app_args,
main_class=args.main_class,
jars=jars,
py_files=py_files,
files=files,
driver_java_options=args.driver_java_options,
driver_library_path=args.driver_library_path,
driver_class_path=args.driver_class_path,
driver_memory=args.driver_memory,
executor_memory=args.executor_memory,
driver_cores=args.driver_cores,
executor_cores=args.executor_cores),
wait=False)
if args.wait:
utils.stream_logs(client=aztk.client,
cluster_id=args.cluster_id,
application_name=args.name)