def execute(args: typing.NamedTuple):
spark_client = aztk.spark.Client(config.load_aztk_screts())
job_conf = JobConfig()
job_conf.merge(args.job_id, args.job_conf)
aztk_applications = []
for application in job_conf.applications:
aztk_applications.append(
aztk.spark.models.ApplicationConfiguration(
name=application.get('name'),
application=application.get('application'),
application_args=application.get('application_args'),
main_class=application.get('main_class'),
jars=[],
py_files=[],
files=[],
driver_java_options=application.get('driver_java_options'),
driver_library_path=application.get('driver_library_path'),
driver_class_path=application.get('driver_class_path'),
driver_memory=application.get('driver_memory'),
executor_memory=application.get('executor_memory'),
driver_cores=application.get('driver_cores'),
executor_cores=application.get('executor_cores')))
# by default, load spark configuration files in .aztk/
spark_configuration = config.load_aztk_spark_config()
# overwrite with values in job_conf if they exist
if job_conf.spark_defaults_conf:
spark_configuration.spark_defaults_conf = job_conf.spark_defaults_conf
if job_conf.spark_env_sh:
spark_configuration.spark_env_sh = job_conf.spark_env_sh
if job_conf.core_site_xml:
spark_configuration.core_site_xml = job_conf.core_site_xml
job_configuration = aztk.spark.models.JobConfiguration(
id=job_conf.id,
applications=aztk_applications,
custom_scripts=job_conf.custom_scripts,
spark_configuration=spark_configuration,
vm_size=job_conf.vm_size,
docker_repo=job_conf.docker_repo,
max_dedicated_nodes=job_conf.max_dedicated_nodes,
max_low_pri_nodes=job_conf.max_low_pri_nodes,
subnet_id=job_conf.subnet_id,
worker_on_master=job_conf.worker_on_master)
#TODO: utils.print_job_conf(job_configuration)
spark_client.submit_job(job_configuration)
def execute(args: typing.NamedTuple):
spark_client = aztk.spark.Client(config.load_aztk_screts())
cluster_conf = ClusterConfiguration()
cluster_conf.spark_configuration = load_aztk_spark_config()
# read cluster.yaml configuartion file, overwrite values with args
file_config, wait = config.read_cluster_config()
cluster_conf.merge(file_config)
cluster_conf.merge(
ClusterConfiguration(cluster_id=args.cluster_id,
vm_count=args.size,
vm_low_pri_count=args.size_low_pri,
vm_size=args.vm_size,
subnet_id=args.subnet_id,
user_configuration=UserConfiguration(
username=args.username,
password=args.password,
),
docker_repo=args.docker_repo))
wait = wait if args.wait is None else args.wait
user_configuration = cluster_conf.user_configuration
if user_configuration and user_configuration.username:
ssh_key, password = utils.get_ssh_key_or_prompt(
spark_client.secrets_config.ssh_pub_key,
user_configuration.username, user_configuration.password,
spark_client.secrets_config)
cluster_conf.user_configuration = aztk.spark.models.UserConfiguration(
username=user_configuration.username,
password=password,
ssh_key=ssh_key)
else:
cluster_conf.user_configuration = None
print_cluster_conf(cluster_conf, wait)
spinner = utils.Spinner()
spinner.start()
# create spark cluster
cluster = spark_client.create_cluster(cluster_conf, wait=wait)
spinner.stop()
if wait:
log.info("Cluster %s created successfully.", cluster.id)
else:
log.info("Cluster %s is being provisioned.", cluster.id)
def execute(args: typing.NamedTuple):
spark_client = aztk.spark.Client(config.load_aztk_screts())
# 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,
subnet_id=args.subnet_id,
wait=args.wait,
username=args.username,
password=args.password,
docker_repo=args.docker_repo)
if cluster_conf.custom_scripts:
custom_scripts = []
for custom_script in cluster_conf.custom_scripts:
custom_scripts.append(
aztk.spark.models.CustomScript(script=custom_script['script'],
run_on=custom_script['runOn']))
else:
custom_scripts = None
if cluster_conf.file_shares:
file_shares = []
for file_share in cluster_conf.file_shares:
file_shares.append(
aztk.spark.models.FileShare(
storage_account_name=file_share['storage_account_name'],
storage_account_key=file_share['storage_account_key'],
file_share_path=file_share['file_share_path'],
mount_path=file_share['mount_path']))
else:
file_shares = None
if cluster_conf.username:
ssh_key, password = utils.get_ssh_key_or_prompt(
spark_client.secrets_config.ssh_pub_key, cluster_conf.username,
cluster_conf.password, spark_client.secrets_config)
user_conf = aztk.spark.models.UserConfiguration(
username=cluster_conf.username, password=password, ssh_key=ssh_key)
else:
user_conf = None
print_cluster_conf(cluster_conf)
spinner = utils.Spinner()
spinner.start()
# create spark cluster
cluster = spark_client.create_cluster(
aztk.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,
subnet_id=cluster_conf.subnet_id,
custom_scripts=custom_scripts,
file_shares=file_shares,
docker_repo=cluster_conf.docker_repo,
spark_configuration=load_aztk_spark_config(),
user_configuration=user_conf),
wait=cluster_conf.wait)
spinner.stop()
if cluster_conf.wait:
log.info("Cluster %s created successfully.", cluster.id)
else:
log.info("Cluster %s is being provisioned.", cluster.id)