batch-shipyard/convoy/storage.py

2373 строки
94 KiB
Python

# Copyright (c) Microsoft Corporation
#
# All rights reserved.
#
# MIT License
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED *AS IS*, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
# DEALINGS IN THE SOFTWARE.
# compat imports
from __future__ import (
absolute_import, division, print_function, unicode_literals
)
from builtins import ( # noqa
bytes, dict, int, list, object, range, str, ascii, chr, hex, input,
next, oct, open, pow, round, super, filter, map, zip)
# stdlib imports
import datetime
import json
import logging
import os
import pickle
import re
import tempfile
import time
import uuid
# non-stdlib imports
import azure.common
import azure.cosmosdb.table as azuretable
import azure.storage.blob as azureblob
import azure.storage.file as azurefile
# local imports
from . import settings
from . import util
# TODO refactor as class
# create logger
logger = logging.getLogger(__name__)
util.setup_logger(logger)
# global defines
_MONITOR_BATCHPOOL_PK = 'BatchPool'
_MONITOR_REMOTEFS_PK = 'RemoteFS'
_ALL_FEDERATIONS_PK = '!!FEDERATIONS'
_FEDERATION_ACTIONS_PREFIX_PK = '!!ACTIONS'
_BLOCKED_FEDERATION_ACTIONS_PREFIX_PK = '!!ACTIONS.BLOCKED'
_MAX_SEQUENCE_ID_PROPERTIES = 15
_MAX_SEQUENCE_IDS_PER_PROPERTY = 975
_DEFAULT_SAS_EXPIRY_DAYS = 365 * 30
_STORAGEACCOUNT = None
_STORAGEACCOUNTKEY = None
_STORAGEACCOUNTEP = None
_STORAGE_CONTAINERS = {
'blob_globalresources': None,
'blob_resourcefiles': None,
'blob_remotefs': None,
'blob_monitoring': None,
'blob_federation_global': None,
'blob_federation': None,
'table_dht': None,
'table_images': None,
'table_globalresources': None,
'table_perf': None,
'table_monitoring': None,
'table_federation_global': None,
'table_federation_jobs': None,
'table_slurm': None,
'queue_federation': None,
# TODO remove following in future release
'table_registry': None,
'blob_torrents': None,
'table_torrentinfo': None,
}
_CONTAINERS_CREATED = set()
def set_storage_configuration(sep, postfix, sa, sakey, saep, sasexpiry):
# type: (str, str, str, str, str, int) -> None
"""Set storage configuration
:param str sep: storage entity prefix
:param str postfix: storage entity postfix
:param str sa: storage account
:param str sakey: storage account key
:param str saep: storage account endpoint
:param int sasexpiry: sas expiry default time in days
"""
if util.is_none_or_empty(sep):
raise ValueError('storage_entity_prefix is invalid')
global _STORAGEACCOUNT, _STORAGEACCOUNTKEY, _STORAGEACCOUNTEP, \
_DEFAULT_SAS_EXPIRY_DAYS
_STORAGE_CONTAINERS['blob_globalresources'] = '-'.join(
(sep + 'gr', postfix))
_STORAGE_CONTAINERS['blob_resourcefiles'] = '-'.join(
(sep + 'rf', postfix))
_STORAGE_CONTAINERS['blob_remotefs'] = sep + 'remotefs'
_STORAGE_CONTAINERS['blob_monitoring'] = sep + 'monitor'
_STORAGE_CONTAINERS['blob_federation'] = sep + 'fed'
_STORAGE_CONTAINERS['blob_federation_global'] = sep + 'fedglobal'
_STORAGE_CONTAINERS['table_dht'] = sep + 'dht'
_STORAGE_CONTAINERS['table_images'] = sep + 'images'
_STORAGE_CONTAINERS['table_globalresources'] = sep + 'gr'
_STORAGE_CONTAINERS['table_perf'] = sep + 'perf'
_STORAGE_CONTAINERS['table_monitoring'] = sep + 'monitor'
_STORAGE_CONTAINERS['table_federation_jobs'] = sep + 'fedjobs'
_STORAGE_CONTAINERS['table_federation_global'] = sep + 'fedglobal'
_STORAGE_CONTAINERS['table_slurm'] = sep + 'slurm'
_STORAGE_CONTAINERS['queue_federation'] = sep + 'fed'
# TODO remove following containers in future release
_STORAGE_CONTAINERS['table_registry'] = sep + 'registry'
_STORAGE_CONTAINERS['blob_torrents'] = '-'.join(
(sep + 'tor', postfix))
_STORAGE_CONTAINERS['table_torrentinfo'] = sep + 'torrentinfo'
# ensure all storage containers are between 3 and 63 chars in length
for key in _STORAGE_CONTAINERS:
length = len(_STORAGE_CONTAINERS[key])
if length < 3 or length > 63:
raise RuntimeError(
'Storage container {} name {} length {} does not fall in '
'storage naming rules. Retry with a modified '
'batch_shipyard:storage_entity_prefix and/or '
'pool_specification:id.'.format(
key, _STORAGE_CONTAINERS[key], length))
_STORAGEACCOUNT = sa
_STORAGEACCOUNTKEY = sakey
_STORAGEACCOUNTEP = saep
if sasexpiry is not None:
_DEFAULT_SAS_EXPIRY_DAYS = sasexpiry
def set_storage_remotefs_container(storage_cluster_id):
# type: (str) -> None
"""Set storage properties for a remotefs storage cluster
:param str storage_cluster_id: storage cluster id
"""
if util.is_none_or_empty(storage_cluster_id):
raise ValueError('storage_cluster_id is invalid')
_STORAGE_CONTAINERS['blob_remotefs'] = '{}-{}'.format(
_STORAGE_CONTAINERS['blob_remotefs'],
storage_cluster_id)
def get_storageaccount():
# type: (None) -> str
"""Get storage account
:rtype: str
:return: storage account
"""
return _STORAGEACCOUNT
def get_storageaccount_key():
# type: (None) -> str
"""Get storage account key
:rtype: str
:return: storage account key
"""
return _STORAGEACCOUNTKEY
def get_storageaccount_endpoint():
# type: (None) -> str
"""Get storage account endpoint
:rtype: str
:return: storage account endpoint
"""
return _STORAGEACCOUNTEP
def get_storage_table_monitoring():
# type: (None) -> str
"""Get the table associated with monitoring
:rtype: str
:return: table name for monitoring
"""
return _STORAGE_CONTAINERS['table_monitoring']
def populate_storage_account_keys_from_aad(storage_mgmt_client, config):
# type: (azure.mgmt.storage.StorageManagementClient, dict) -> None
"""Fetch secrets with secret ids in config from keyvault
:param azure.mgmt.storage.StorageManagementClient storage_mgmt_client:
storage client
:param dict config: configuration dict
"""
modified = False
if storage_mgmt_client is None:
return modified
# get ref to main storage account
main_sa = settings.batch_shipyard_settings(config).storage_account_settings
# iterate all storage accounts, if storage account does not have
# a storage account key, then lookup via aad
for ssel in settings.iterate_storage_credentials(config):
sc = settings.credentials_storage(config, ssel)
if util.is_none_or_empty(sc.account_key):
if util.is_none_or_empty(sc.resource_group):
raise ValueError(
('resource_group is invalid for storage account {} to '
'be retrieved by aad').format(sc.account))
keys = storage_mgmt_client.storage_accounts.list_keys(
sc.resource_group, sc.account)
props = storage_mgmt_client.storage_accounts.get_properties(
sc.resource_group, sc.account)
if main_sa == ssel:
ep = props.primary_endpoints.table
if ep is None:
raise ValueError(
('the specified '
'batch_shipyard:storage_account_settings storage '
'account (link={} account={}) is not a general '
'purpose type storage account').format(
ssel, sc.account))
else:
# get either blob or file endpoint
ep = (
props.primary_endpoints.blob or
props.primary_endpoints.file
)
ep = '.'.join(ep.rstrip('/').split('.')[2:])
settings.set_credentials_storage_account(
config, ssel, keys.keys[0].value, ep)
modified = True
return modified
def generate_blob_container_uri(storage_settings, container):
# type: (StorageCredentialsSettings, str) -> str
"""Create a uri to a blob container
:param StorageCredentialsSettings storage_settings: storage settings
:param str container: container
:rtype: str
:return: blob container uri
"""
blob_client = azureblob.BlockBlobService(
account_name=storage_settings.account,
account_key=storage_settings.account_key,
endpoint_suffix=storage_settings.endpoint)
return '{}://{}/{}'.format(
blob_client.protocol, blob_client.primary_endpoint, container)
def create_blob_container_saskey(
storage_settings, container, kind, create_container=False):
# type: (StorageCredentialsSettings, str, str, bool) -> str
"""Create a saskey for a blob container
:param StorageCredentialsSettings storage_settings: storage settings
:param str container: container
:param str kind: ingress or egress
:param bool create_container: create container
:rtype: str
:return: saskey
"""
global _CONTAINERS_CREATED
blob_client = azureblob.BlockBlobService(
account_name=storage_settings.account,
account_key=storage_settings.account_key,
endpoint_suffix=storage_settings.endpoint)
if create_container:
key = 'blob:{}:{}:{}'.format(
storage_settings.account, storage_settings.endpoint, container)
if key not in _CONTAINERS_CREATED:
blob_client.create_container(container, fail_on_exist=False)
_CONTAINERS_CREATED.add(key)
if kind == 'ingress':
perm = azureblob.ContainerPermissions(read=True, list=True)
elif kind == 'egress':
perm = azureblob.ContainerPermissions(
read=True, write=True, delete=True, list=True)
else:
raise ValueError('{} type of transfer not supported'.format(kind))
return blob_client.generate_container_shared_access_signature(
container, perm,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=_DEFAULT_SAS_EXPIRY_DAYS)
)
def create_file_share_saskey(
storage_settings, file_share, kind, create_share=False):
# type: (StorageCredentialSettings, str, str, bool) -> str
"""Create a saskey for a file share
:param StorageCredentialsSettings storage_settings: storage settings
:param str file_share: file share
:param str kind: ingress or egress
:param bool create_share: create file share
:rtype: str
:return: saskey
"""
file_client = azurefile.FileService(
account_name=storage_settings.account,
account_key=storage_settings.account_key,
endpoint_suffix=storage_settings.endpoint)
if create_share:
key = 'file:{}:{}:{}'.format(
storage_settings.account, storage_settings.endpoint, file_share)
if key not in _CONTAINERS_CREATED:
file_client.create_share(file_share, fail_on_exist=False)
_CONTAINERS_CREATED.add(key)
if kind == 'ingress':
perm = azurefile.SharePermissions(read=True, list=True)
elif kind == 'egress':
perm = azurefile.SharePermissions(
read=True, write=True, delete=True, list=True)
else:
raise ValueError('{} type of transfer not supported'.format(kind))
return file_client.generate_share_shared_access_signature(
file_share, perm,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=_DEFAULT_SAS_EXPIRY_DAYS)
)
def create_saskey(
storage_settings, path, file, create, list_perm, read, write, delete,
expiry_days=None):
# type: (settings.StorageCredentialsSettings, str, bool, bool, bool, bool,
# bool, bool, int) -> None
"""Create an object-level sas key
:param settings.StorageCredentialsSetting storage_settings:
storage settings
:param str path: path
:param bool file: file sas
:param bool create: create perm
:param bool list_perm: list perm
:param bool read: read perm
:param bool write: write perm
:param bool delete: delete perm
:param int expiry_days: expiry in days
:rtype: str
:return: sas token
"""
if expiry_days is None:
expiry_days = _DEFAULT_SAS_EXPIRY_DAYS
if file:
client = azurefile.FileService(
account_name=storage_settings.account,
account_key=storage_settings.account_key,
endpoint_suffix=storage_settings.endpoint)
tmp = path.split('/')
if len(tmp) < 1:
raise ValueError('path is invalid: {}'.format(path))
share_name = tmp[0]
if len(tmp) == 1:
perm = azurefile.SharePermissions(
read=read, write=write, delete=delete, list=list_perm)
sas = client.generate_share_shared_access_signature(
share_name=share_name, permission=perm,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=expiry_days)
)
else:
if len(tmp) == 2:
directory_name = ''
file_name = tmp[1]
else:
directory_name = tmp[1]
file_name = '/'.join(tmp[2:])
perm = azurefile.FilePermissions(
read=read, create=create, write=write, delete=delete)
sas = client.generate_file_shared_access_signature(
share_name=share_name, directory_name=directory_name,
file_name=file_name, permission=perm,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=expiry_days)
)
else:
client = azureblob.BlockBlobService(
account_name=storage_settings.account,
account_key=storage_settings.account_key,
endpoint_suffix=storage_settings.endpoint)
tmp = path.split('/')
if len(tmp) < 1:
raise ValueError('path is invalid: {}'.format(path))
container_name = tmp[0]
if len(tmp) == 1:
perm = azureblob.ContainerPermissions(
read=read, write=write, delete=delete, list=list_perm)
sas = client.generate_container_shared_access_signature(
container_name=container_name, permission=perm,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=expiry_days)
)
else:
blob_name = '/'.join(tmp[1:])
perm = azureblob.BlobPermissions(
read=read, create=create, write=write, delete=delete)
sas = client.generate_blob_shared_access_signature(
container_name=container_name, blob_name=blob_name,
permission=perm,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=expiry_days)
)
return sas
def _construct_partition_key_from_config(config, pool_id=None):
# type: (dict, str) -> str
"""Construct partition key from config
:param dict config: configuration dict
:param str pool_id: use specified pool id instead
:rtype: str
:return: partition key
"""
if util.is_none_or_empty(pool_id):
pool_id = settings.pool_id(config, lower=True)
return '{}${}'.format(
settings.credentials_batch(config).account, pool_id)
def _add_global_resource(
blob_client, table_client, config, pk, dr, grtype):
# type: (azureblob.BlockBlobService, azuretable.TableService, dict, str,
# settings.DataReplicationSettings, str) -> None
"""Add global resources
:param azure.storage.blob.BlockService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str pk: partition key
:param settings.DataReplicationSettings dr: data replication settings
:param str grtype: global resources type
"""
try:
if grtype == 'docker_images':
prefix = 'docker'
resources = settings.global_resources_docker_images(config)
elif grtype == 'singularity_images':
prefix = 'singularity'
resources = settings.global_resources_singularity_images(config)
key_fingerprint_dict = (
settings.singularity_signed_images_key_fingerprint_dict(
config))
else:
raise NotImplementedError(
'global resource type: {}'.format(grtype))
for gr in resources:
resource = '{}:{}'.format(prefix, gr)
resource_sha1 = util.hash_string(resource)
logger.info('adding global resource: {} hash={}'.format(
resource, resource_sha1))
key_fingerprint = None
if prefix == 'singularity':
key_fingerprint = key_fingerprint_dict.get(gr, None)
table_client.insert_or_replace_entity(
_STORAGE_CONTAINERS['table_globalresources'],
{
'PartitionKey': pk,
'RowKey': resource_sha1,
'Resource': resource,
'KeyFingerprint': key_fingerprint,
}
)
for i in range(0, dr.concurrent_source_downloads):
blob_client.create_blob_from_bytes(
container_name=_STORAGE_CONTAINERS['blob_globalresources'],
blob_name='{}.{}'.format(resource_sha1, i),
blob=b'',
)
except KeyError:
pass
def populate_global_resource_blobs(blob_client, table_client, config):
# type: (azureblob.BlockBlobService, azuretable.TableService, dict) -> None
"""Populate global resource blobs
:param azure.storage.blob.BlockService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
"""
pk = _construct_partition_key_from_config(config)
dr = settings.data_replication_settings(config)
_add_global_resource(
blob_client, table_client, config, pk, dr, 'docker_images')
_add_global_resource(
blob_client, table_client, config, pk, dr, 'singularity_images')
def add_resources_to_monitor(table_client, config, pools, fsmap):
# type: (azuretable.TableService, dict, List[str], dict) -> None
"""Populate resources to monitor
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param list pools: pools to monitor
:param dict fsmap: fs clusters to monitor
"""
if util.is_not_empty(pools):
bc = settings.credentials_batch(config)
for poolid in pools:
entity = {
'PartitionKey': _MONITOR_BATCHPOOL_PK,
'RowKey': '{}${}'.format(bc.account, poolid),
'BatchServiceUrl': bc.account_service_url,
'AadEndpoint': bc.aad.endpoint,
'AadAuthorityUrl': bc.aad.authority_url,
}
if settings.verbose(config):
logger.debug(
'inserting pool monitor resource entity: {}'.format(
entity))
try:
table_client.insert_entity(
_STORAGE_CONTAINERS['table_monitoring'], entity)
except azure.common.AzureConflictHttpError:
logger.error('monitoring for pool {} already exists'.format(
poolid))
else:
logger.debug('resource monitor added for pool {}'.format(
poolid))
if util.is_not_empty(fsmap):
for sc_id in fsmap:
fs = fsmap[sc_id]
entity = {
'PartitionKey': _MONITOR_REMOTEFS_PK,
'RowKey': sc_id,
'Type': fs['type'],
'ResourceGroup': fs['rg'],
'NodeExporterPort': fs['ne_port'],
'VMs': json.dumps(fs['vms'], ensure_ascii=False),
}
if fs['type'] == 'glusterfs':
entity['AvailabilitySet'] = fs['as']
if settings.verbose(config):
logger.debug(
'inserting RemoteFS monitor resource entity: {}'.format(
entity))
try:
table_client.insert_entity(
_STORAGE_CONTAINERS['table_monitoring'], entity)
except azure.common.AzureConflictHttpError:
logger.error(
'monitoring for remotefs {} already exists'.format(sc_id))
else:
logger.debug('resource monitor added for remotefs {}'.format(
sc_id))
def list_monitored_resources(table_client, config):
# type: (azuretable.TableService, dict) -> None
"""List monitored resources
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
"""
# list batch pools monitored
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_monitoring'],
filter='PartitionKey eq \'{}\''.format(_MONITOR_BATCHPOOL_PK))
except azure.common.AzureMissingResourceHttpError:
logger.error(
'cannot list monitored Batch pools as monitoring table does '
'not exist')
else:
pools = ['batch pools monitored:']
for ent in entities:
ba, poolid = ent['RowKey'].split('$')
pools.append('* pool id: {} (account: {})'.format(
poolid, ba))
if len(pools) == 1:
logger.info('no Batch pools monitored')
else:
logger.info('{}'.format('\n'.join(pools)))
del pools
# list remotefs monitored
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_monitoring'],
filter='PartitionKey eq \'{}\''.format(_MONITOR_REMOTEFS_PK))
except azure.common.AzureMissingResourceHttpError:
logger.error(
'cannot list monitored RemoteFS clusters as monitoring table does '
'not exist')
else:
fs = ['RemoteFS clusters monitored:']
for ent in entities:
sc_id = ent['RowKey']
fs.append('* storage cluster id: {}'.format(sc_id))
fs.append(' * type: {}'.format(ent['Type']))
fs.append(' * resource group: {}'.format(ent['ResourceGroup']))
if len(fs) == 1:
logger.info('no RemoteFS clusters monitored')
else:
logger.info('{}'.format('\n'.join(fs)))
def remove_resources_from_monitoring(
table_client, config, all, pools, fsclusters):
# type: (azuretable.TableService, dict, bool, List[str], List[str]) -> None
"""Remove resources from monitoring
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param bool all: all resource monitors
:param list pools: pools to remove from monitoring
:param list fsclusters: fs clusters to remove from monitoring
"""
if all:
if util.confirm_action(
config, 'remove all resources from monitoring'):
_clear_table(
table_client, _STORAGE_CONTAINERS['table_monitoring'], config,
pool_id=None, pk=_MONITOR_BATCHPOOL_PK)
return
if util.is_not_empty(pools):
bc = settings.credentials_batch(config)
for poolid in pools:
try:
table_client.delete_entity(
_STORAGE_CONTAINERS['table_monitoring'],
partition_key=_MONITOR_BATCHPOOL_PK,
row_key='{}${}'.format(bc.account, poolid)
)
except azure.common.AzureMissingResourceHttpError:
logger.error('pool {} is not monitored'.format(poolid))
else:
logger.debug('resource monitor removed for pool {}'.format(
poolid))
if util.is_not_empty(fsclusters):
for sc_id in fsclusters:
try:
table_client.delete_entity(
_STORAGE_CONTAINERS['table_monitoring'],
partition_key=_MONITOR_REMOTEFS_PK,
row_key=sc_id
)
except azure.common.AzureMissingResourceHttpError:
logger.error('RemoteFS cluster {} is not monitored'.format(
sc_id))
else:
logger.debug(
'resource monitor removed for RemoteFS cluster {}'.format(
sc_id))
def hash_pool_and_service_url(pool_id, batch_service_url):
"""Hash a pool and service url
:param str pool_id: pool id
:param str batch_service_url: batch_service_url
:rtype: str
:return: hashed pool and service url
"""
return util.hash_string('{}${}'.format(
batch_service_url.rstrip('/'), pool_id))
def hash_federation_id(federation_id):
"""Hash a federation id
:param str federation_id: federation id
:rtype: str
:return: hashed federation id
"""
fedhash = util.hash_string(federation_id)
logger.debug('federation id {} -> {}'.format(federation_id, fedhash))
return fedhash
def generate_job_id_locator_partition_key(federation_id, job_id):
"""Hash a job id locator
:param str federation_id: federation id
:param str job_id: job id
:rtype: str
:return: hashed fedhash and job id
"""
return '{}${}'.format(
util.hash_string(federation_id), util.hash_string(job_id))
def create_federation_id(
blob_client, table_client, queue_client, config, federation_id, force,
unique_jobs):
# type: (azure.storage.blob.BlockBlobService, azuretable.TableService,
# azure.queue.QueueService, dict, str, bool, bool) -> None
"""Create storage containers for federation id
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param azure.storage.queue.QueueService queue_service: queue client
:param dict config: configuration dict
:param str federation_id: federation id
:param bool force: force creation
:param bool unique_jobs: unique job ids required
"""
fedhash = hash_federation_id(federation_id)
# create table entry for federation id
entity = {
'PartitionKey': _ALL_FEDERATIONS_PK,
'RowKey': fedhash,
'FederationId': federation_id,
'BatchShipyardFederationVersion': 1,
'UniqueJobIds': unique_jobs,
}
logger.debug(
'inserting federation {} entity to global table '
'(unique_jobs={})'.format(federation_id, unique_jobs))
try:
table_client.insert_entity(
_STORAGE_CONTAINERS['table_federation_global'], entity)
except azure.common.AzureConflictHttpError:
logger.error('federation id {} already exists'.format(
federation_id))
if force:
if util.confirm_action(
config, 'overwrite existing federation {}; this can '
'result in undefined behavior'.format(federation_id)):
table_client.insert_or_replace_entity(
_STORAGE_CONTAINERS['table_federation_global'], entity)
else:
return
else:
return
# create blob container for federation id
contname = '{}-{}'.format(_STORAGE_CONTAINERS['blob_federation'], fedhash)
logger.debug('creating container: {}'.format(contname))
blob_client.create_container(contname)
# create job queue for federation id
queuename = '{}-{}'.format(
_STORAGE_CONTAINERS['queue_federation'], fedhash)
logger.debug('creating queue: {}'.format(queuename))
queue_client.create_queue(queuename)
if settings.raw(config):
rawout = {
'federation': {
'id': entity['FederationId'],
'hash': entity['RowKey'],
'batch_shipyard_federation_version':
entity['BatchShipyardFederationVersion'],
'unique_job_ids': entity['UniqueJobIds'],
'storage': {
'account': get_storageaccount(),
'endpoint': get_storageaccount_endpoint(),
'containers': {
'queue': queuename,
'blob': contname,
},
},
}
}
print(json.dumps(rawout, sort_keys=True, indent=4))
def federation_requires_unique_job_ids(table_client, federation_id):
fedhash = hash_federation_id(federation_id)
try:
entity = table_client.get_entity(
_STORAGE_CONTAINERS['table_federation_global'],
_ALL_FEDERATIONS_PK, fedhash)
except azure.common.AzureMissingResourceHttpError:
raise RuntimeError(
'federation {} does not exist'.format(federation_id))
return entity['UniqueJobIds']
def list_federations(table_client, config, federation_ids):
# type: (azuretable.TableService, dict, List[str]) -> None
"""List all federations
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str federation_id: federation id
"""
log = []
if util.is_not_empty(federation_ids):
log.append('listing federations: {}'.format(', '.join(federation_ids)))
fedhashset = set()
fedhashmap = {}
for x in federation_ids:
fid = x.lower()
fhash = hash_federation_id(fid)
fedhashmap[fhash] = fid
fedhashset.add(fhash)
else:
log.append('listing all federations:')
fedhashset = None
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_global'],
filter='PartitionKey eq \'{}\''.format(_ALL_FEDERATIONS_PK))
except azure.common.AzureMissingResourceHttpError:
logger.error('no federations exist')
return
if settings.raw(config):
rawout = {}
for ent in entities:
fedhash = ent['RowKey']
if fedhashset is not None and fedhash not in fedhashset:
continue
if settings.raw(config):
rawout[ent['FederationId']] = {
'hash': fedhash,
'batch_shipyard_federation_version':
ent['BatchShipyardFederationVersion'],
'unique_job_ids': ent['UniqueJobIds'],
'pools': {}
}
else:
log.append('* federation id: {}'.format(ent['FederationId']))
log.append(' * federation hash: {}'.format(fedhash))
log.append(' * batch shipyard federation version: {}'.format(
ent['BatchShipyardFederationVersion']))
log.append(' * unique job ids: {}'.format(ent['UniqueJobIds']))
log.append(' * pools:')
# get list of pools associated with federation
try:
fedents = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_global'],
filter='PartitionKey eq \'{}\''.format(fedhash))
except azure.common.AzureMissingResourceHttpError:
continue
numpools = 0
for fe in fedents:
numpools += 1
if settings.raw(config):
rawout[ent['FederationId']]['pools'][fe['PoolId']] = {
'batch_account': fe['BatchAccount'],
'location': fe['Location'],
'hash': fe['RowKey'],
}
else:
log.append(' * pool id: {}'.format(fe['PoolId']))
log.append(' * batch account: {}'.format(
fe['BatchAccount']))
log.append(' * location: {}'.format(fe['Location']))
log.append(' * pool hash: {}'.format(fe['RowKey']))
if numpools == 0:
log.append(' * no pools in federation')
# get number of jobs/job schedules for federation
_, fejobs = get_all_federation_jobs(table_client, fedhash)
fejobs = list(fejobs)
fejk = [x['Kind'] for x in fejobs]
if settings.raw(config):
rawout[ent['FederationId']]['num_jobs'] = fejk.count('job')
rawout[ent['FederationId']]['num_job_schedules'] = fejk.count(
'job_schedule')
else:
log.append(' * number of jobs: {}'.format(fejk.count('job')))
log.append(' * number of job schedules: {}'.format(
fejk.count('job_schedule')))
if settings.raw(config):
print(json.dumps(rawout, sort_keys=True, indent=4))
else:
if len(log) > 1:
logger.info(os.linesep.join(log))
else:
logger.error('no federations exist')
def batch_delete_entities(table_client, table_name, pk, rks):
if util.is_none_or_empty(rks):
return
i = 0
tb = azuretable.TableBatch()
for rk in rks:
tb.delete_entity(pk, rk)
i += 1
if i == 100:
table_client.commit_batch(table_name, tb)
tb = azuretable.TableBatch()
i = 0
if i > 0:
table_client.commit_batch(table_name, tb)
def collate_all_location_entities_for_job(table_client, fedhash, entity):
loc_pk = '{}${}'.format(fedhash, entity['RowKey'])
rks = []
try:
loc_entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}\''.format(loc_pk))
except azure.common.AzureMissingResourceHttpError:
pass
else:
for loc_entity in loc_entities:
rks.append(loc_entity['RowKey'])
return loc_pk, rks
def get_all_federation_jobs(table_client, fedhash):
pk = '{}${}'.format(_FEDERATION_ACTIONS_PREFIX_PK, fedhash)
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}\''.format(pk))
except azure.common.AzureMissingResourceHttpError:
entities = []
return pk, entities
def gc_federation_jobs(table_client, config, federation_id, fedhash):
# retrieve all job sequence rows for federation
pk, entities = get_all_federation_jobs(table_client, fedhash)
gc_dict = {
pk: []
}
# process all jobs
for entity in entities:
# if sequence exists, ask for confirmation
if ('Sequence0' in entity and
util.is_not_empty(entity['Sequence0']) and
not util.confirm_action(
config,
msg=('destroying pending actions for job {} in '
'federation id {}').format(
entity['RowKey'], federation_id))):
raise RuntimeError(
'Not destroying federation job {} with pending actions '
'in federation id {}'.format(
entity['RowKey'], federation_id))
gc_dict[pk].append(entity['RowKey'])
loc_pk, loc_rks = collate_all_location_entities_for_job(
table_client, fedhash, entity)
if util.is_not_empty(loc_rks) and not util.confirm_action(
config,
msg='orphan job {} in federation id {}'.format(
entity['RowKey'], federation_id)):
raise RuntimeError(
'Not orphaning active/completed federation job '
'{} in federation id {}'.format(
entity['RowKey'], federation_id))
gc_dict[loc_pk] = loc_rks
# batch delete entities
for gc_pk in gc_dict:
batch_delete_entities(
table_client, _STORAGE_CONTAINERS['table_federation_jobs'],
gc_pk, gc_dict[gc_pk])
def destroy_federation_id(
blob_client, table_client, queue_client, config, federation_id):
# type: (azure.storage.blob.BlockBlobService, azuretable.TableService,
# azure.queue.QueueService, dict, str) -> None
"""Remove storage containers for federation id
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param azure.storage.queue.QueueService queue_service: queue client
:param dict config: configuration dict
:param str federation_id: federation id
"""
fedhash = hash_federation_id(federation_id)
# delete table entities for federation id
logger.debug('deleting all federation {} job entities'.format(
federation_id))
gc_federation_jobs(table_client, config, federation_id, fedhash)
# remove table entry for federation id
logger.debug('deleting federation {} entities in global table'.format(
federation_id))
try:
table_client.delete_entity(
_STORAGE_CONTAINERS['table_federation_global'],
_ALL_FEDERATIONS_PK, fedhash)
except azure.common.AzureMissingResourceHttpError:
pass
try:
fedentities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_global'],
filter='PartitionKey eq \'{}\''.format(fedhash))
except azure.common.AzureMissingResourceHttpError:
pass
else:
batch_delete_entities(
table_client, _STORAGE_CONTAINERS['table_federation_global'],
fedhash, [x['RowKey'] for x in fedentities])
# delete job queue for federation id
queuename = '{}-{}'.format(
_STORAGE_CONTAINERS['queue_federation'], fedhash)
logger.debug('deleting queue: {}'.format(queuename))
queue_client.delete_queue(queuename)
# delete blob container for federation id
contname = '{}-{}'.format(_STORAGE_CONTAINERS['blob_federation'], fedhash)
logger.debug('deleting container: {}'.format(contname))
blob_client.delete_container(contname)
if settings.raw(config):
rawout = {
'federation': {
'id': federation_id,
'hash': fedhash,
'storage': {
'account': get_storageaccount(),
'endpoint': get_storageaccount_endpoint(),
'containers': {
'queue': queuename,
'blob': contname,
},
},
},
}
print(json.dumps(rawout, sort_keys=True, indent=4))
def _check_if_federation_exists(table_client, fedhash):
try:
table_client.get_entity(
_STORAGE_CONTAINERS['table_federation_global'],
_ALL_FEDERATIONS_PK, fedhash)
except azure.common.AzureMissingResourceHttpError:
return False
return True
def add_pool_to_federation(
table_client, config, federation_id, batch_service_url, pools):
# type: (azuretable.TableService, dict, str, str, List[str]) -> None
"""Populate federation with pools
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str federation_id: federation id
:param str batch_service_url: batch service url to associate
:param list pools: pools to monitor
"""
fedhash = hash_federation_id(federation_id)
# check if federation exists
if not _check_if_federation_exists(table_client, fedhash):
logger.error('federation {} does not exist'.format(federation_id))
return
if util.is_not_empty(batch_service_url):
batch_service_url = batch_service_url.rstrip('/')
account, location = settings.parse_batch_service_url(
batch_service_url)
else:
bc = settings.credentials_batch(config)
batch_service_url = bc.account_service_url.rstrip('/')
account, location = settings.parse_batch_service_url(
batch_service_url)
rawout = {
'federation': {
'id': federation_id,
'hash': fedhash,
'storage': {
'account': get_storageaccount(),
'endpoint': get_storageaccount_endpoint(),
},
},
'pools_added': {}
}
for poolid in pools:
rk = hash_pool_and_service_url(poolid, batch_service_url)
entity = {
'PartitionKey': fedhash,
'RowKey': rk,
'FederationId': federation_id,
'BatchAccount': account,
'PoolId': poolid,
'Location': location,
'BatchServiceUrl': batch_service_url,
}
if settings.verbose(config):
logger.debug(
'inserting pool federation entity: {}'.format(
entity))
try:
table_client.insert_entity(
_STORAGE_CONTAINERS['table_federation_global'], entity)
except azure.common.AzureConflictHttpError:
logger.error(
'federation {} entity for pool {} already exists'.format(
federation_id, poolid))
else:
logger.debug('federation {} entity added for pool {}'.format(
federation_id, poolid))
if settings.raw(config):
rawout['pools_added'][entity['RowKey']] = {
'pool_id': entity['PoolId'],
'batch_account': entity['BatchAccount'],
'location': entity['Location'],
'batch_service_url': entity['BatchServiceUrl'],
}
if settings.raw(config) and util.is_not_empty(rawout['pools_added']):
print(json.dumps(rawout, sort_keys=True, indent=4))
def remove_pool_from_federation(
table_client, config, federation_id, all, batch_service_url, pools):
# type: (azuretable.TableService, dict, str, bool, str, List[str]) -> None
"""Remove pools from federation
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str federation_id: federation id
:param bool all: all pools
:param str batch_service_url: batch service url to associate
:param list pools: pools to monitor
"""
fedhash = hash_federation_id(federation_id)
# check if federation exists
if not _check_if_federation_exists(table_client, fedhash):
logger.error('federation {} does not exist'.format(federation_id))
return
rawout = {
'federation': {
'id': federation_id,
'hash': fedhash,
'storage': {
'account': get_storageaccount(),
'endpoint': get_storageaccount_endpoint(),
},
},
'pools_removed': {}
}
logger.warning(
'**WARNING** Removing active pools with jobs/job schedules in a '
'federation can lead to orphaned data. It is recommended to delete '
'all federation jobs/job schedules associated with the pools '
'to be removed prior to removal from the federation!')
if all:
if util.confirm_action(
config, 'remove all pools from federation {}'.format(
federation_id)):
if settings.raw(config):
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_global'],
filter='PartitionKey eq \'{}\''.format(fedhash))
except azure.common.AzureMissingResourceHttpError:
pass
else:
for entity in entities:
rawout['pools_removed'][entity['RowKey']] = {
'hash': entity['PoolId'],
'batch_account': entity['BatchAccount'],
'location': entity['Location'],
'batch_service_url': entity['BatchServiceUrl'],
}
_clear_table(
table_client, _STORAGE_CONTAINERS['table_federation_global'],
config, pool_id=None, pk=fedhash)
if (settings.raw(config) and
util.is_not_empty(rawout['pools_removed'])):
print(json.dumps(rawout, sort_keys=True, indent=4))
return
if util.is_not_empty(batch_service_url):
account, _ = settings.parse_batch_service_url(batch_service_url)
else:
bc = settings.credentials_batch(config)
batch_service_url = bc.account_service_url
account, _ = settings.parse_batch_service_url(batch_service_url)
for poolid in pools:
if not util.confirm_action(
config, 'remove pool {} from federation {}'.format(
poolid, federation_id)):
continue
try:
rk = hash_pool_and_service_url(poolid, batch_service_url)
entity = None
if settings.raw(config):
entity = table_client.get_entity(
_STORAGE_CONTAINERS['table_federation_global'],
partition_key=fedhash,
row_key=rk,
)
table_client.delete_entity(
_STORAGE_CONTAINERS['table_federation_global'],
partition_key=fedhash,
row_key=rk,
)
except azure.common.AzureMissingResourceHttpError:
logger.error('pool {} is not in federation {}'.format(
poolid, federation_id))
else:
logger.debug('pool {} removed from federation {}'.format(
poolid, federation_id))
if settings.raw(config):
rawout['pools_removed'][entity['RowKey']] = {
'pool_id': entity['PoolId'],
'batch_account': entity['BatchAccount'],
'location': entity['Location'],
'batch_service_url': entity['BatchServiceUrl'],
}
if settings.raw(config) and util.is_not_empty(rawout['pools_removed']):
print(json.dumps(rawout, sort_keys=True, indent=4))
def _pack_sequences(ent, unique_id):
seq = []
for i in range(0, _MAX_SEQUENCE_ID_PROPERTIES):
prop = 'Sequence{}'.format(i)
if prop in ent and util.is_not_empty(ent[prop]):
seq.extend(ent[prop].split(','))
seq.append(str(unique_id))
if len(seq) > _MAX_SEQUENCE_IDS_PER_PROPERTY * _MAX_SEQUENCE_ID_PROPERTIES:
raise RuntimeError(
'maximum number of enqueued sequence ids reached, please allow '
'job actions to drain')
for i in range(0, _MAX_SEQUENCE_ID_PROPERTIES):
prop = 'Sequence{}'.format(i)
start = i * _MAX_SEQUENCE_IDS_PER_PROPERTY
end = start + _MAX_SEQUENCE_IDS_PER_PROPERTY
if end > len(seq):
end = len(seq)
if start < end:
ent[prop] = ','.join(seq[start:end])
else:
ent[prop] = None
def _retrieve_and_merge_sequence(
table_client, pk, unique_id, kind, target, entity_must_not_exist):
rk = util.hash_string(target)
try:
ent = table_client.get_entity(
_STORAGE_CONTAINERS['table_federation_jobs'], pk, rk)
if entity_must_not_exist:
raise RuntimeError(
'{} {} action entity already exists: rolling back action '
'due to unique job id requirement for federation.'.format(
kind, target))
except azure.common.AzureMissingResourceHttpError:
ent = {
'PartitionKey': pk,
'RowKey': rk,
'Kind': kind,
'Id': target,
}
_pack_sequences(ent, unique_id)
return ent
def _insert_or_merge_entity_with_etag(table_client, table_name, entity):
if 'etag' not in entity:
try:
table_client.insert_entity(table_name, entity=entity)
return True
except azure.common.AzureConflictHttpError:
pass
else:
etag = entity['etag']
entity.pop('etag')
try:
table_client.merge_entity(table_name, entity=entity, if_match=etag)
return True
except azure.common.AzureConflictHttpError:
pass
except azure.common.AzureHttpError as ex:
if ex.status_code != 412:
raise
return False
def check_if_job_exists_in_federation(
table_client, federation_id, job_id):
pk = generate_job_id_locator_partition_key(federation_id, job_id)
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}\''.format(pk))
for ent in entities:
return True
except azure.common.AzureMissingResourceHttpError:
pass
return False
def check_if_job_is_terminated_in_federation(
table_client, federation_id, job_id):
pk = generate_job_id_locator_partition_key(federation_id, job_id)
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}\''.format(pk))
except azure.common.AzureMissingResourceHttpError:
return False
else:
for ent in entities:
if 'TerminateTimestamp' in ent:
return True
return False
def add_job_to_federation(
table_client, queue_client, config, federation_id, unique_id, msg,
kind):
# type: (azure.cosmosdb.TableService, azure.queue.QueueService, str,
# uuid.UUID, dict, str) -> None
"""Add a job/job schedule to a federation
:param azure.cosmosdb.table.TableService table_client: table client
:param azure.storage.queue.QueueService queue_service: queue client
:param str federation_id: federation id
:param uuid.UUID unique_id: unique id
:param dict msg: dict payload
:param str kind: kind
"""
requires_unique_job_ids = federation_requires_unique_job_ids(
table_client, federation_id)
fedhash = hash_federation_id(federation_id)
pk = '{}${}'.format(_FEDERATION_ACTIONS_PREFIX_PK, fedhash)
target = msg['target']
# check if job is terminated first
if check_if_job_is_terminated_in_federation(
table_client, federation_id, target):
if requires_unique_job_ids:
raise RuntimeError(
'cannot add {} {} as federation requires unique job '
'ids'.format(kind, target))
if not util.confirm_action(
config,
'adding {} although one or more {}s representing this {} '
'in federation {} have been terminated'.format(
target, kind, kind, federation_id)):
raise RuntimeError(
'aborted adding {} {} to federation {}'.format(
kind, target, federation_id))
# upsert unique id to sequence
while True:
entity = _retrieve_and_merge_sequence(
table_client, pk, unique_id, kind, target, requires_unique_job_ids)
if _insert_or_merge_entity_with_etag(
table_client, _STORAGE_CONTAINERS['table_federation_jobs'],
entity):
logger.debug(
'upserted {} {} sequence uid {} to federation {}'.format(
kind, target, unique_id, federation_id))
break
else:
logger.debug(
'conflict upserting {} {} sequence uid {} to '
'federation {}'.format(kind, target, unique_id, federation_id))
# add queue message
msg_data = json.dumps(msg, ensure_ascii=True, sort_keys=True)
contname = '{}-{}'.format(
_STORAGE_CONTAINERS['queue_federation'], fedhash)
queue_client.put_message(contname, msg_data, time_to_live=-1)
def list_blocked_actions_in_federation(
table_client, config, federation_id, job_id, job_schedule_id):
# type: (azure.cosmosdb.TableService, dict, str, str, str) -> None
"""List blocked actions in federation
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str federation_id: federation id
:param str job_id: job id
:param str job_schedule_id: job schedule id
"""
fedhash = hash_federation_id(federation_id)
pk = '{}${}'.format(_BLOCKED_FEDERATION_ACTIONS_PREFIX_PK, fedhash)
if (util.is_none_or_empty(job_id) and
util.is_none_or_empty(job_schedule_id)):
logger.debug(
'fetching all blocked jobs/job schedules for federation '
'id {}'.format(federation_id))
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}\''.format(pk))
except azure.common.AzureMissingResourceHttpError:
pass
else:
rk = util.hash_string(
job_id if util.is_not_empty(job_id) else job_schedule_id)
try:
entities = [table_client.get_entity(
_STORAGE_CONTAINERS['table_federation_jobs'], pk, rk)]
except azure.common.AzureMissingResourceHttpError:
pass
if settings.raw(config):
log = {}
else:
log = [
'listing blocked jobs/job schedules for federation id {}:'.format(
federation_id)
]
for entity in entities:
id = entity['Id']
if settings.raw(config):
log[id] = {
'hash': entity['RowKey'],
'unique_id': entity['UniqueId'],
'task_group_size': entity['NumTasks'],
'reason': entity['Reason'],
}
else:
log.append('* id: {}'.format(id))
log.append(' * hash: {}'.format(entity['RowKey']))
log.append(' * unique id: {}'.format(entity['UniqueId']))
log.append(' * task group size: {}'.format(entity['NumTasks']))
log.append(' * reason: {}'.format(entity['Reason']))
if settings.raw(config):
print(json.dumps(log, sort_keys=True, indent=4))
else:
if len(log) > 1:
logger.info(os.linesep.join(log))
else:
logger.debug(
'no blocked jobs/job schedules exist in federation '
'id {}'.format(federation_id))
def list_queued_actions_in_federation(
table_client, config, federation_id, job_id, job_schedule_id):
# type: (azure.cosmosdb.TableService, dict, str, str, str) -> None
"""List queued actions in federation
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str federation_id: federation id
:param str job_id: job id
:param str job_schedule_id: job schedule id
"""
fedhash = hash_federation_id(federation_id)
pk = '{}${}'.format(_FEDERATION_ACTIONS_PREFIX_PK, fedhash)
if (util.is_none_or_empty(job_id) and
util.is_none_or_empty(job_schedule_id)):
logger.debug(
'fetching all queued jobs/job schedules for federation '
'id {}'.format(federation_id))
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}\''.format(pk))
except azure.common.AzureMissingResourceHttpError:
pass
else:
rk = util.hash_string(
job_id if util.is_not_empty(job_id) else job_schedule_id)
try:
entities = [table_client.get_entity(
_STORAGE_CONTAINERS['table_federation_jobs'], pk, rk)]
except azure.common.AzureMissingResourceHttpError:
pass
if settings.raw(config):
log = {}
else:
log = [
'listing queued jobs/job schedules for federation id {}:'.format(
federation_id)
]
for entity in entities:
if ('Sequence0' not in entity or
util.is_none_or_empty(entity['Sequence0'])):
continue
id = entity['Id']
uids = entity['Sequence0'].split(',')[:10]
if settings.raw(config):
log[id] = {
'kind': entity['Kind'],
'hash': entity['RowKey'],
'first_ten_unique_ids': uids,
}
else:
log.append('* id: {}'.format(id))
log.append(' * kind: {}'.format(entity['Kind']))
log.append(' * hash: {}'.format(entity['RowKey']))
log.append(' * first ten unique ids:')
for uid in uids:
log.append(' * {}'.format(uid))
if settings.raw(config):
print(json.dumps(log, sort_keys=True, indent=4))
else:
if len(log) > 1:
logger.info(os.linesep.join(log))
else:
logger.debug(
'no queued jobs/job schedules exist in federation '
'id {}'.format(federation_id))
def list_active_jobs_in_federation(
table_client, config, federation_id, job_id, job_schedule_id):
# type: (azure.cosmosdb.TableService, dict, str, str, str) -> None
"""List active jobs in federation
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str federation_id: federation id
:param str job_id: job id
:param str job_schedule_id: job schedule id
"""
fedhash = hash_federation_id(federation_id)
if (util.is_none_or_empty(job_id) and
util.is_none_or_empty(job_schedule_id)):
targets = []
logger.debug(
'fetching all active jobs/job schedules for federation '
'id {}'.format(federation_id))
_, entities = get_all_federation_jobs(table_client, fedhash)
for entity in entities:
targets.append(entity['RowKey'])
else:
targets = [job_id if util.is_not_empty(job_id) else job_schedule_id]
if len(targets) == 0:
logger.error(
'no active jobs/job schedules in federation id {}'.format(
federation_id))
return
if settings.raw(config):
log = {}
else:
log = [
'listing active jobs/job schedules for federation id {}:'.format(
federation_id)
]
for targethash in targets:
try:
entities = table_client.query_entities(
_STORAGE_CONTAINERS['table_federation_jobs'],
filter='PartitionKey eq \'{}${}\''.format(fedhash, targethash))
except azure.common.AzureMissingResourceHttpError:
pass
kind = None
for ent in entities:
id = ent['Id']
if kind is None:
kind = ent['Kind']
if settings.raw(config):
log[id] = {
'type': kind,
'hash': targethash,
}
else:
log.append('* id: {}'.format(id))
log.append(' * type: {}'.format(kind))
log.append(' * hash: {}'.format(targethash))
if 'AdditionTimestamps' in ent:
ats = ent['AdditionTimestamps'].split(',')[-10:]
else:
ats = None
if 'UniqueIds' in ent:
uids = ent['UniqueIds'].split(',')[-10:]
else:
uids = None
if settings.raw(config):
poolid = ent['PoolId']
log[id][poolid] = {
'batch_account': ent['BatchAccount'],
'service_url': ent['ServiceUrl'],
}
log[id][poolid]['ten_most_recent_task_additions'] = ats
log[id][poolid]['ten_most_recent_unique_ids_serviced'] = uids
log[id][poolid]['terminate_timestamp'] = (
ent['TerminateTimestamp'] if 'TerminateTimestamp' in ent
else None
)
else:
log.append(' * pool id: {}'.format(ent['PoolId']))
log.append(' * batch account: {}'.format(
ent['BatchAccount']))
log.append(' * service url: {}'.format(ent['ServiceUrl']))
log.append(' * ten most recent task addition times:')
if util.is_not_empty(ats):
for at in ats:
log.append(' * {}'.format(at))
else:
log.append(' * n/a')
log.append(' * ten most recent unique ids serviced:')
if util.is_not_empty(uids):
for uid in uids:
log.append(' * {}'.format(uid))
else:
log.append(' * n/a')
log.append(' * termination time: {}'.format(
ent['TerminateTimestamp'] if 'TerminateTimestamp' in ent
else 'n/a'))
if settings.raw(config):
print(json.dumps(log, sort_keys=True, indent=4))
else:
if len(log) > 1:
logger.info(os.linesep.join(log))
else:
logger.error(
'no active jobs/job schedules exist in federation '
'id {}'.format(federation_id))
def pickle_and_upload(blob_client, data, rpath, federation_id=None):
# type: (azureblob.BlockBlobService, dict, str, str) -> str
"""Pickle and upload data to a given remote path
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param dict data: data to pickle
:param str rpath: remote path
:param str federation_id: federation id
:rtype: str
:return: sas url of uploaded pickle
"""
f = tempfile.NamedTemporaryFile(mode='wb', delete=False)
fname = f.name
try:
with open(fname, 'wb') as f:
pickle.dump(data, f, protocol=pickle.HIGHEST_PROTOCOL)
f.close()
if util.is_none_or_empty(federation_id):
sas_urls = upload_resource_files(blob_client, [(rpath, fname)])
else:
sas_urls = upload_job_for_federation(
blob_client, federation_id, [(rpath, fname)])
if len(sas_urls) != 1:
raise RuntimeError(
'unexpected number of sas urls for pickled upload')
return next(iter(sas_urls.values()))
finally:
try:
os.unlink(fname)
except OSError:
pass
del f
del fname
def delete_or_terminate_job_from_federation(
blob_client, table_client, queue_client, config, delete, federation_id,
job_id, job_schedule_id, all_jobs, all_jobschedules, force):
# type: (azure.storage.blob.BlockBlobService, azure.cosmosdb.TableService,
# azure.queue.QueueService, bool, str, str, str, bool,
# bool, bool) -> None
"""Delete or terminate a job from a federation
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param azure.storage.queue.QueueService queue_service: queue client
:param bool delete: delete instead of terminate
:param str federation_id: federation id
:param str job_id: job id
:param str job_schedule_id: job schedule id
:param bool all_jobs: all jobs
:param bool all_jobschedules: all jobschedules
:param bool force: force
"""
fedhash = hash_federation_id(federation_id)
if all_jobs or all_jobschedules:
if all_jobs:
kind = 'job'
elif all_jobschedules:
kind = 'job_schedule'
targets = []
logger.debug('fetching all {}s for federation id {}'.format(
kind, federation_id))
pk, entities = get_all_federation_jobs(table_client, fedhash)
for entity in entities:
if entity['Kind'] == kind:
targets.append(entity['RowKey'])
else:
pk = '{}${}'.format(_FEDERATION_ACTIONS_PREFIX_PK, fedhash)
kind = 'job' if util.is_not_empty(job_id) else 'job_schedule'
targets = job_id if util.is_not_empty(job_id) else job_schedule_id
method = 'delete' if delete else 'terminate'
if len(targets) == 0:
logger.error(
'no {}s to {} in federation id {}'.format(
kind, method, federation_id))
return
raw_output = {}
for target in targets:
# if terminate, check if job exists
if not force and method == 'terminate':
if not check_if_job_exists_in_federation(
table_client, federation_id, target):
logger.warning(
'skipping termination of non-existent job {} in '
'federation {}'.format(target, federation_id))
continue
if not util.confirm_action(
config,
msg='{} {} id {} in federation {}'.format(
method, kind, target, federation_id)):
return
unique_id = uuid.uuid4()
rpath = 'messages/{}.pickle'.format(unique_id)
# upload message data to blob
info = {
'version': '1',
'action': {
'method': method,
'kind': kind,
},
kind: {
'id': target,
},
}
sas_url = pickle_and_upload(
blob_client, info, rpath, federation_id=federation_id)
# upsert unique id to sequence
while True:
entity = _retrieve_and_merge_sequence(
table_client, pk, unique_id, kind, target, False)
if _insert_or_merge_entity_with_etag(
table_client, _STORAGE_CONTAINERS['table_federation_jobs'],
entity):
logger.debug(
'upserted {} {} sequence uid {} to federation {}'.format(
kind, target, unique_id, federation_id))
break
else:
logger.debug(
'conflict upserting {} {} sequence uid {} to '
'federation {}'.format(
kind, target, unique_id, federation_id))
# add queue message
msg = {
'version': '1',
'federation_id': federation_id,
'target': target,
'blob_data': sas_url,
'uuid': str(unique_id),
}
msg_data = json.dumps(msg, ensure_ascii=True, sort_keys=True)
contname = '{}-{}'.format(
_STORAGE_CONTAINERS['queue_federation'], fedhash)
queue_client.put_message(contname, msg_data, time_to_live=-1)
logger.debug('enqueued {} of {} {} for federation {}'.format(
method, kind, target, federation_id))
if settings.raw(config):
raw_output[target] = {
'federation': {
'id': federation_id,
'storage': {
'account': get_storageaccount(),
'endpoint': get_storageaccount_endpoint(),
},
},
'kind': kind,
'action': method,
'unique_id': str(unique_id),
}
if util.is_not_empty(raw_output):
print(json.dumps(raw_output, indent=4, sort_keys=True))
def zap_unique_id_from_federation(
blob_client, config, federation_id, unique_id):
# type: (azure.storage.blob.BlockBlobService, dict, str, str) -> None
"""Zap a unique id from a federation
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param dict config: configuration dict
:param str federation_id: federation id
:param str unique_id: unique id
"""
jloc = 'messages/{}.pickle'.format(unique_id)
deleted = delete_resource_file(
blob_client, jloc, federation_id=federation_id)
if deleted and settings.raw(config):
rawout = {
'federation': {
'id': federation_id,
'storage': {
'account': get_storageaccount(),
'endpoint': get_storageaccount_endpoint(),
},
},
'unique_id': unique_id,
}
print(json.dumps(rawout, sort_keys=True, indent=4))
def create_slurm_partition(
table_client, queue_client, config, cluster_id, partition_name,
batch_service_url, pool_id, compute_node_type, max_compute_nodes,
hostlist):
partpool_hash = util.hash_string('{}-{}'.format(
partition_name, batch_service_url, pool_id))
# insert partition entity
entity = {
'PartitionKey': 'PARTITIONS${}'.format(cluster_id),
'RowKey': '{}${}'.format(partition_name, partpool_hash),
'BatchServiceUrl': batch_service_url,
'BatchPoolId': pool_id,
'ComputeNodeType': compute_node_type,
'HostList': hostlist,
'BatchShipyardSlurmVersion': 1,
}
logger.debug(
'inserting slurm partition {}:{} entity to table for '
'cluster {}'.format(partition_name, pool_id, cluster_id))
try:
table_client.insert_entity(_STORAGE_CONTAINERS['table_slurm'], entity)
except azure.common.AzureConflictHttpError:
logger.error('partition {}:{} cluster id {} already exists'.format(
partition_name, pool_id, cluster_id))
if util.confirm_action(
config, 'overwrite existing partition {}:{} for '
'cluster {}; this can result in undefined behavior'.format(
partition_name, pool_id, cluster_id)):
table_client.insert_or_replace_entity(
_STORAGE_CONTAINERS['table_slurm'], entity)
else:
raise
# create queue
qname = '{}-{}'.format(cluster_id, partpool_hash)
logger.debug('creating queue: {}'.format(qname))
queue_client.create_queue(qname)
def get_slurm_host_node_id(table_client, cluster_id, host):
node_id = None
try:
entity = table_client.get_entity(
_STORAGE_CONTAINERS['table_slurm'],
'{}${}'.format('HOSTS', cluster_id), host)
node_id = entity['BatchNodeId']
except (azure.common.AzureMissingResourceHttpError, KeyError):
pass
return node_id
def clear_slurm_table_entities(table_client, cluster_id):
logger.debug('deleting slurm cluster {} entities in table'.format(
cluster_id))
tablename = _STORAGE_CONTAINERS['table_slurm']
keys = ['HOSTS', 'PARTITIONS']
for key in keys:
try:
pk = '{}${}'.format(key, cluster_id)
entities = table_client.query_entities(
tablename,
filter='PartitionKey eq \'{}\''.format(pk))
except azure.common.AzureMissingResourceHttpError:
pass
else:
batch_delete_entities(
table_client, tablename, pk, [x['RowKey'] for x in entities]
)
def _check_file_and_upload(blob_client, file, key, container=None):
# type: (azure.storage.blob.BlockBlobService, tuple, str, str) -> None
"""Upload file to blob storage if necessary
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param tuple file: file to upload
:param str key: blob container key
:param str container: absolute container override
"""
if file[0] is None:
return
contname = container or _STORAGE_CONTAINERS[key]
upload = True
# check if blob exists
try:
prop = blob_client.get_blob_properties(contname, file[0])
if (prop.properties.content_settings.content_md5 ==
util.compute_md5_for_file(file[1], True)):
logger.debug(
'remote file is the same for {}, skipping'.format(file[0]))
upload = False
except azure.common.AzureMissingResourceHttpError:
pass
if upload:
logger.info('uploading file {} as {!r}'.format(file[1], file[0]))
blob_client.create_blob_from_path(contname, file[0], str(file[1]))
def delete_resource_file(blob_client, blob_name, federation_id=None):
# type: (azure.storage.blob.BlockBlobService, str) -> bool
"""Delete a resource file from blob storage
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param str blob_name: blob name
:param str federation_id: federation id
"""
if util.is_not_empty(federation_id):
fedhash = hash_federation_id(federation_id)
container = '{}-{}'.format(
_STORAGE_CONTAINERS['blob_federation'], fedhash)
else:
container = _STORAGE_CONTAINERS['blob_resourcefiles']
try:
blob_client.delete_blob(container, blob_name)
logger.debug('blob {} deleted from container {}'.format(
blob_name, container))
except azure.common.AzureMissingResourceHttpError:
logger.warning('blob {} does not exist in container {}'.format(
blob_name, container))
return False
return True
def upload_resource_files(blob_client, files):
# type: (azure.storage.blob.BlockBlobService, List[tuple]) -> dict
"""Upload resource files to blob storage
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param list files: files to upload
:rtype: dict
:return: sas url dict
"""
sas_urls = {}
for file in files:
_check_file_and_upload(blob_client, file, 'blob_resourcefiles')
sas_urls[file[0]] = 'https://{}.blob.{}/{}/{}?{}'.format(
_STORAGEACCOUNT, _STORAGEACCOUNTEP,
_STORAGE_CONTAINERS['blob_resourcefiles'], file[0],
blob_client.generate_blob_shared_access_signature(
_STORAGE_CONTAINERS['blob_resourcefiles'], file[0],
permission=azureblob.BlobPermissions.READ,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=_DEFAULT_SAS_EXPIRY_DAYS)
)
)
return sas_urls
def upload_for_nonbatch(blob_client, files, kind):
# type: (azure.storage.blob.BlockBlobService, List[tuple],
# str) -> List[str]
"""Upload files to blob storage for non-batch
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param dict config: configuration dict
:param list files: files to upload
:param str kind: "remotefs", "monitoring" or "federation"
:rtype: list
:return: list of file urls
"""
if kind == 'federation':
kind = '{}_global'.format(kind.lower())
key = 'blob_{}'.format(kind.lower())
ret = []
for file in files:
_check_file_and_upload(blob_client, file, key)
ret.append('https://{}.blob.{}/{}/{}'.format(
_STORAGEACCOUNT, _STORAGEACCOUNTEP,
_STORAGE_CONTAINERS[key], file[0]))
return ret
def upload_to_container(blob_client, sa, files, container, gen_sas=True):
# type: (azure.storage.blob.BlockBlobService,
# settings.StorageCredentialsSettings, List[tuple],
# str, bool) -> dict
"""Upload files to a specific blob storage container
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param settings.StorageCredentialsSettings sa: storage account
:param list files: files to upload
:param str container: container
:param bool gen_sas: generate a SAS URL for blob
:rtype: dict
:return: sas url dict
"""
sas_urls = {}
for file in files:
_check_file_and_upload(blob_client, file, None, container=container)
sas_urls[file[0]] = 'https://{}.blob.{}/{}/{}'.format(
sa.account, sa.endpoint, container, file[0],
)
if gen_sas:
sas_urls[file[0]] = '{}?{}'.format(
sas_urls[file[0]],
blob_client.generate_blob_shared_access_signature(
container, file[0],
permission=azureblob.BlobPermissions.READ,
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=_DEFAULT_SAS_EXPIRY_DAYS)
)
)
return sas_urls
def create_global_lock_blob(blob_client, kind):
# type: (azure.storage.blob.BlockBlobService, str) -> None
"""Create a global lock blob
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param str kind: "remotefs", "monitoring" or "federation"
"""
if kind == 'federation':
kind = '{}_global'.format(kind.lower())
key = 'blob_{}'.format(kind.lower())
blob_client.create_blob_from_bytes(
_STORAGE_CONTAINERS[key], 'global.lock', b'')
def upload_job_for_federation(blob_client, federation_id, files):
# type: (azure.storage.blob.BlockBlobService, str,
# List[tuple]) -> List[str]
"""Upload files to blob storage for federation jobs
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param str federation_id: federation id
:param list files: files to upload
:rtype: list
:return: list of file urls
"""
fedhash = hash_federation_id(federation_id)
contname = '{}-{}'.format(_STORAGE_CONTAINERS['blob_federation'], fedhash)
sas_urls = {}
for file in files:
_check_file_and_upload(blob_client, file, None, container=contname)
sas_urls[file[0]] = 'https://{}.blob.{}/{}/{}?{}'.format(
_STORAGEACCOUNT, _STORAGEACCOUNTEP,
contname, file[0],
blob_client.generate_blob_shared_access_signature(
contname, file[0],
permission=azureblob.BlobPermissions(read=True, delete=True),
expiry=datetime.datetime.utcnow() +
datetime.timedelta(days=_DEFAULT_SAS_EXPIRY_DAYS)
)
)
return sas_urls
def delete_storage_containers(
blob_client, table_client, config, skip_tables=False):
# type: (azureblob.BlockBlobService, azuretable.TableService,
# dict, bool) -> None
"""Delete storage containers
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param bool skip_tables: skip deleting tables
"""
for key in _STORAGE_CONTAINERS:
if key == 'table_registry' or key == 'table_torrentinfo':
# TODO remove in future release: unused table
logger.debug('deleting table: {}'.format(_STORAGE_CONTAINERS[key]))
table_client.delete_table(_STORAGE_CONTAINERS[key])
elif key == 'blob_torrents':
# TODO remove in future release: unused container
logger.debug('deleting container: {}'
.format(_STORAGE_CONTAINERS[key]))
blob_client.delete_container(_STORAGE_CONTAINERS[key])
elif key.startswith('blob_'):
if (key == 'blob_remotefs' or key == 'blob_monitoring' or
key == 'blob_federation' or
key == 'blob_federation_global'):
continue
logger.debug('deleting container: {}'.format(
_STORAGE_CONTAINERS[key]))
blob_client.delete_container(_STORAGE_CONTAINERS[key])
elif not skip_tables and key.startswith('table_'):
logger.debug('deleting table: {}'.format(_STORAGE_CONTAINERS[key]))
table_client.delete_table(_STORAGE_CONTAINERS[key])
def _clear_blobs(blob_client, container):
# type: (azureblob.BlockBlobService, str) -> None
"""Clear blobs in container
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param str container: container to clear blobs from
"""
logger.info('deleting blobs: {}'.format(container))
try:
blobs = blob_client.list_blobs(container)
except azure.common.AzureMissingResourceHttpError:
logger.warning('container not found: {}'.format(container))
else:
for blob in blobs:
blob_client.delete_blob(container, blob.name)
def _clear_blob_task_resourcefiles(blob_client, container, config):
# type: (azureblob.BlockBlobService, str, dict) -> None
"""Clear task resource file blobs in container
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param str container: container to clear blobs from
:param dict config: configuration dict
"""
bs = settings.batch_shipyard_settings(config)
envfileloc = '{}taskrf-'.format(bs.storage_entity_prefix)
logger.info('deleting blobs with prefix: {}'.format(envfileloc))
try:
blobs = blob_client.list_blobs(container, prefix=envfileloc)
except azure.common.AzureMissingResourceHttpError:
logger.warning('container not found: {}'.format(container))
else:
for blob in blobs:
blob_client.delete_blob(container, blob.name)
def _clear_table(table_client, table_name, config, pool_id=None, pk=None):
# type: (azuretable.TableService, str, dict, str, str) -> None
"""Clear table entities
:param azure.cosmosdb.table.TableService table_client: table client
:param str table_name: table name
:param dict config: configuration dict
:param str pool_id: use specified pool id instead
:param str pk: partition key
"""
if pk is None:
pk = _construct_partition_key_from_config(config, pool_id=pool_id)
logger.debug('clearing table (pk={}): {}'.format(pk, table_name))
ents = table_client.query_entities(
table_name, filter='PartitionKey eq \'{}\''.format(pk))
# batch delete entities
i = 0
bet = azuretable.TableBatch()
for ent in ents:
bet.delete_entity(ent['PartitionKey'], ent['RowKey'])
i += 1
if i == 100:
table_client.commit_batch(table_name, bet)
bet = azuretable.TableBatch()
i = 0
if i > 0:
table_client.commit_batch(table_name, bet)
def clear_storage_containers(
blob_client, table_client, config, tables_only=False, pool_id=None):
# type: (azureblob.BlockBlobService, azuretable.TableService, dict,
# bool, str) -> None
"""Clear storage containers
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param bool tables_only: clear only tables
:param str pool_id: use specified pool id instead
"""
bs = settings.batch_shipyard_settings(config)
for key in _STORAGE_CONTAINERS:
# TODO remove in a future release: unused container
if key == 'blob_torrents':
continue
if not tables_only and key.startswith('blob_'):
if (key == 'blob_remotefs' or key == 'blob_monitoring' or
key == 'blob_federation' or
key == 'blob_federation_global'):
continue
_clear_blobs(blob_client, _STORAGE_CONTAINERS[key])
elif key.startswith('table_'):
# TODO remove in a future release: unused table
if key == 'table_registry' or key == 'table_torrentinfo':
continue
if (key == 'table_monitoring' or
key == 'table_federation_global' or
key == 'table_federation_jobs' or
key == 'table_slurm'):
continue
try:
_clear_table(
table_client, _STORAGE_CONTAINERS[key], config,
pool_id=pool_id)
except azure.common.AzureMissingResourceHttpError:
if key != 'table_perf' or bs.store_timing_metrics:
raise
def delete_or_clear_diagnostics_logs(blob_client, config, delete):
# type: (azureblob.BlockBlobService, dict, bool) -> None
"""Clear diagnostics logs container
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param dict config: configuration dict
:param bool delete: delete instead of clear
"""
bs = settings.batch_shipyard_settings(config)
cont = bs.storage_entity_prefix + '-diaglogs'
if not util.confirm_action(
config, '{} diagnostics logs'.format(
'delete' if delete else 'clear')):
return
if delete:
logger.debug('deleting container: {}'.format(cont))
blob_client.delete_container(cont)
else:
_clear_blobs(blob_client, cont)
def create_storage_containers(blob_client, table_client, config):
# type: (azureblob.BlockBlobService, azuretable.TableService, dict) -> None
"""Create storage containers
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
"""
bs = settings.batch_shipyard_settings(config)
for key in _STORAGE_CONTAINERS:
if key.startswith('blob_'):
# TODO remove in a future release: unused container
if key == 'blob_torrents':
continue
if (key == 'blob_remotefs' or key == 'blob_monitoring' or
key == 'blob_federation' or
key == 'blob_federation_global'):
continue
logger.info('creating container: {}'.format(
_STORAGE_CONTAINERS[key]))
while True:
blob_client.create_container(_STORAGE_CONTAINERS[key])
if blob_client.exists(_STORAGE_CONTAINERS[key]):
break
time.sleep(1)
elif key.startswith('table_'):
# TODO remove in a future release: unused table
if key == 'table_registry' or key == 'table_torrentinfo':
continue
if (key == 'table_monitoring' or
key == 'table_federation_global' or
key == 'table_federation_jobs'):
continue
if key == 'table_perf' and not bs.store_timing_metrics:
continue
logger.info('creating table: {}'.format(_STORAGE_CONTAINERS[key]))
while True:
table_client.create_table(_STORAGE_CONTAINERS[key])
if table_client.exists(_STORAGE_CONTAINERS[key]):
break
time.sleep(1)
def create_storage_containers_nonbatch(
blob_client, table_client, queue_client, kind):
# type: (azureblob.BlockBlobService, azuretable.TableService,
# azurequeue.QueueService, str) -> None
"""Create storage containers used for non-batch actions
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param azure.storage.queue.QueueService queue_service: queue client
:param str kind: kind, "remotefs", "monitoring", "federation", or "slurm"
"""
if kind == 'federation':
create_storage_containers_nonbatch(
blob_client, table_client, None, 'federation_global')
create_storage_containers_nonbatch(
None, table_client, None, 'federation_jobs')
else:
if blob_client is not None:
try:
key = 'blob_{}'.format(kind.lower())
contname = _STORAGE_CONTAINERS[key]
except KeyError:
pass
else:
logger.info('creating container: {}'.format(contname))
while True:
blob_client.create_container(contname)
if blob_client.exists(contname):
break
time.sleep(1)
if table_client is not None:
try:
key = 'table_{}'.format(kind.lower())
contname = _STORAGE_CONTAINERS[key]
except KeyError:
pass
else:
logger.info('creating table: {}'.format(contname))
while True:
table_client.create_table(contname)
if table_client.exists(contname):
break
time.sleep(1)
if queue_client is not None:
try:
key = 'queue_{}'.format(kind.lower())
contname = _STORAGE_CONTAINERS[key]
except KeyError:
pass
else:
logger.info('creating queue: {}'.format(contname))
while True:
queue_client.create_queue(contname)
if queue_client.exists(contname):
break
time.sleep(1)
def delete_storage_containers_nonbatch(
blob_client, table_client, queue_client, kind):
# type: (azureblob.BlockBlobService, azuretable.TableService,
# azurequeue.QueueService, str) -> None
"""Delete storage containers used for non-batch actions
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param azure.storage.queue.QueueService queue_service: queue client
:param str kind: kind, "remotefs", "monitoring" or "federation"
"""
if kind == 'federation':
delete_storage_containers_nonbatch(
blob_client, table_client, queue_client, 'federation_global')
delete_storage_containers_nonbatch(
None, table_client, None, 'federation_jobs')
else:
if blob_client is not None:
try:
key = 'blob_{}'.format(kind.lower())
contname = _STORAGE_CONTAINERS[key]
except KeyError:
pass
else:
logger.info('deleting container: {}'.format(contname))
try:
blob_client.delete_container(contname)
except azure.common.AzureMissingResourceHttpError:
logger.warning('container not found: {}'.format(contname))
if table_client is not None:
try:
key = 'table_{}'.format(kind.lower())
contname = _STORAGE_CONTAINERS[key]
except KeyError:
pass
else:
logger.debug('deleting table: {}'.format(contname))
try:
table_client.delete_table(contname)
except azure.common.AzureMissingResourceHttpError:
logger.warning('table not found: {}'.format(contname))
if queue_client is not None:
try:
key = 'queue_{}'.format(kind.lower())
contname = _STORAGE_CONTAINERS[key]
except KeyError:
pass
else:
logger.debug('deleting queue: {}'.format(contname))
try:
queue_client.delete_queue(contname)
except azure.common.AzureMissingResourceHttpError:
logger.warning('queue not found: {}'.format(contname))
def delete_file_share_directory(storage_settings, share, directory):
# type: (StorageCredentialsSettings, str, str) -> None
"""Delete file share directory recursively
:param StorageCredentialsSettings storage_settings: storage settings
:param str share: share
:param str directory: directory to delete
"""
file_client = azurefile.FileService(
account_name=storage_settings.account,
account_key=storage_settings.account_key,
endpoint_suffix=storage_settings.endpoint)
logger.info(
'recursively deleting files and directories in share {} at '
'directory {}'.format(share, directory))
del_dirs = []
dirs = [directory]
while len(dirs) > 0:
dir = dirs.pop()
try:
objects = file_client.list_directories_and_files(
share, directory_name=dir)
except azure.common.AzureMissingResourceHttpError:
logger.warning('directory {} does not exist on share {}'.format(
directory, share))
continue
del_dirs.append(dir)
for obj in objects:
path = '{}/{}'.format(dir or '', obj.name)
if type(obj) == azurefile.models.File:
logger.debug('deleting file {} on share {}'.format(
path, share))
file_client.delete_file(share, '', path)
else:
dirs.append(path)
del_dirs.append(path)
for dir in del_dirs[::-1]:
logger.debug('deleting directory {} on share {}'.format(dir, share))
file_client.delete_directory(share, dir)
def delete_storage_containers_boot_diagnostics(
blob_client, vm_name, vm_id):
# type: (azureblob.BlockBlobService, str, str) -> None
"""Delete storage containers used for remotefs bootdiagnostics
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param str vm_name: vm name
:param str vm_id: vm id
"""
name = re.sub('[\W_]+', '', vm_name) # noqa
contname = 'bootdiagnostics-{}-{}'.format(
name[0:min((9, len(name)))], vm_id)
logger.info('deleting container: {}'.format(contname))
try:
blob_client.delete_container(contname)
except azure.common.AzureMissingResourceHttpError:
logger.warning('container not found: {}'.format(contname))
def cleanup_with_del_pool(blob_client, table_client, config, pool_id=None):
# type: (azureblob.BlockBlobService, azuretable.TableService,
# dict, str) -> None
"""Special cleanup routine in combination with delete pool
:param azure.storage.blob.BlockBlobService blob_client: blob client
:param azure.cosmosdb.table.TableService table_client: table client
:param dict config: configuration dict
:param str pool_id: pool id
"""
if util.is_none_or_empty(pool_id):
pool_id = settings.pool_id(config)
if not util.confirm_action(
config, 'delete/cleanup of Batch Shipyard metadata in storage '
'containers associated with {} pool'.format(pool_id)):
return
clear_storage_containers(
blob_client, table_client, config, tables_only=True, pool_id=pool_id)
delete_storage_containers(
blob_client, table_client, config, skip_tables=True)