def delete_cluster_instances(cluster_name, region=None, force=None):
"""
Initiate the forced termination of all cluster compute nodes. Does not work with AWS Batch clusters.
:param cluster_name: Name of the cluster
:type cluster_name: str
:param region: AWS Region that the operation corresponds to.
:type region: str
:param force: Force the deletion also when the cluster with the given name is not found. (Defaults to 'false'.)
:type force: bool
:rtype: None
"""
cluster = Cluster(cluster_name)
try:
if not check_cluster_version(cluster):
raise BadRequestException(
f"Cluster '{cluster_name}' belongs to an incompatible ParallelCluster major version."
)
if cluster.stack.scheduler == "awsbatch":
raise BadRequestException(
"the delete cluster instances operation does not support AWS Batch clusters."
)
except StackNotFoundError:
if not force:
raise NotFoundException(
f"Cluster '{cluster_name}' does not exist or belongs to an incompatible ParallelCluster major version. "
"To force the deletion of all compute nodes, please use the `force` param."
)
cluster.terminate_nodes()
def validate_cluster(cluster: Cluster):
try:
if not check_cluster_version(cluster):
raise BadRequestException(
f"Cluster '{cluster.name}' belongs to an incompatible ParallelCluster major version."
)
except StackNotFoundError:
raise NotFoundException(
f"Cluster '{cluster.name}' does not exist or belongs to an incompatible ParallelCluster major version."
)
def _get_underlying_image_or_stack(imagebuilder):
image = None
stack = None
try:
image = imagebuilder.image
except NonExistingImageError:
try:
stack = imagebuilder.stack
except NonExistingStackError:
raise NotFoundException(
f"No image or stack associated with ParallelCluster image id: {imagebuilder.image_id}."
)
return image, stack
def wrapper(*args, **kwargs):
try:
return func(*args, **kwargs)
except ParallelClusterApiException as e:
raise e
except (LimitExceeded, LimitExceededError) as e:
raise LimitExceededException(str(e)) from e
except (BadRequest, BadRequestError) as e:
raise BadRequestException(str(e)) from e
except Conflict as e:
raise ConflictException(str(e)) from e
except NotFound as e:
raise NotFoundException(str(e)) from e
except Exception as e:
raise InternalServiceException(str(e)) from e
def describe_image(image_id, region=None):
"""
Get detailed information about an existing image.
:param image_id: Id of the image
:type image_id: str
:param region: AWS Region that the operation corresponds to.
:type region: str
:rtype: DescribeImageResponseContent
"""
imagebuilder = ImageBuilder(image_id=image_id)
try:
return _image_to_describe_image_response(imagebuilder)
except NonExistingImageError:
try:
return _stack_to_describe_image_response(imagebuilder)
except NonExistingStackError:
raise NotFoundException("No image or stack associated with ParallelCluster image id: {}.".format(image_id))
def delete_cluster(cluster_name, region=None):
"""
Initiate the deletion of a cluster.
:param cluster_name: Name of the cluster
:type cluster_name: str
:param region: AWS Region that the operation corresponds to.
:type region: str
:rtype: DeleteClusterResponseContent
"""
try:
cluster = Cluster(cluster_name)
if not check_cluster_version(cluster):
raise BadRequestException(
f"Cluster '{cluster_name}' belongs to an incompatible ParallelCluster major version."
)
if not cluster.status == CloudFormationStackStatus.DELETE_IN_PROGRESS:
# TODO: remove keep_logs logic from delete
cluster.delete(keep_logs=False)
return DeleteClusterResponseContent(
cluster=ClusterInfoSummary(
cluster_name=cluster_name,
cloudformation_stack_status=CloudFormationStackStatus.DELETE_IN_PROGRESS,
cloudformation_stack_arn=cluster.stack.id,
region=os.environ.get("AWS_DEFAULT_REGION"),
version=cluster.stack.version,
cluster_status=cloud_formation_status_to_cluster_status(CloudFormationStackStatus.DELETE_IN_PROGRESS),
)
)
except StackNotFoundError:
raise NotFoundException(
f"Cluster '{cluster_name}' does not exist or belongs to an incompatible ParallelCluster major version. "
"In case you have running instances belonging to a deleted cluster please use the DeleteClusterInstances "
"API."
)
def update_cluster(
update_cluster_request_content: Dict,
cluster_name,
suppress_validators=None,
validation_failure_level=None,
region=None,
dryrun=None,
force_update=None,
):
"""
Update a cluster managed in a given region.
:param update_cluster_request_content:
:param cluster_name: Name of the cluster
:type cluster_name: str
:param suppress_validators: Identifies one or more config validators to suppress.
Format: (ALL|type:[A-Za-z0-9]+)
:type suppress_validators: List[str]
:param validation_failure_level: Min validation level that will cause the update to fail.
(Defaults to 'error'.)
:type validation_failure_level: dict | bytes
:param region: AWS Region that the operation corresponds to.
:type region: str
:param dryrun: Only perform request validation without creating any resource.
May be used to validate the cluster configuration and update requirements. Response code: 200
:type dryrun: bool
:param force_update: Force update by ignoring the update validation errors.
(Defaults to 'false'.)
:type force_update: bool
:rtype: UpdateClusterResponseContent
"""
# Set defaults
validation_failure_level = validation_failure_level or ValidationLevel.ERROR
dryrun = dryrun is True
force_update = force_update is True
update_cluster_request_content = UpdateClusterRequestContent.from_dict(update_cluster_request_content)
cluster_config = update_cluster_request_content.cluster_configuration
if not cluster_config:
LOGGER.error("Failed: configuration is required and cannot be empty")
raise BadRequestException("configuration is required and cannot be empty")
try:
cluster = Cluster(cluster_name)
if not check_cluster_version(cluster, exact_match=True):
raise BadRequestException(
f"the update can be performed only with the same ParallelCluster version ({cluster.stack.version}) "
"used to create the cluster."
)
if dryrun:
_, changes, ignored_validation_failures = cluster.validate_update_request(
target_source_config=cluster_config,
force=force_update,
validator_suppressors=get_validator_suppressors(suppress_validators),
validation_failure_level=FailureLevel[validation_failure_level],
)
change_set, _ = _analyze_changes(changes)
validation_messages = validation_results_to_config_validation_errors(ignored_validation_failures)
raise DryrunOperationException(change_set=change_set, validation_messages=validation_messages or None)
changes, ignored_validation_failures = cluster.update(
target_source_config=cluster_config,
validator_suppressors=get_validator_suppressors(suppress_validators),
validation_failure_level=FailureLevel[validation_failure_level],
force=force_update,
)
change_set, _ = _analyze_changes(changes)
return UpdateClusterResponseContent(
cluster=ClusterInfoSummary(
cluster_name=cluster_name,
cloudformation_stack_status=CloudFormationStackStatus.UPDATE_IN_PROGRESS,
cloudformation_stack_arn=cluster.stack.id,
region=os.environ.get("AWS_DEFAULT_REGION"),
version=cluster.stack.version,
cluster_status=cloud_formation_status_to_cluster_status(CloudFormationStackStatus.UPDATE_IN_PROGRESS),
),
validation_messages=validation_results_to_config_validation_errors(ignored_validation_failures) or None,
change_set=change_set,
)
except ConfigValidationError as e:
config_validation_messages = validation_results_to_config_validation_errors(e.validation_failures) or None
raise UpdateClusterBadRequestException(
UpdateClusterBadRequestExceptionResponseContent(
configuration_validation_errors=config_validation_messages, message=str(e)
)
)
except ClusterUpdateError as e:
raise _handle_cluster_update_error(e)
except (NotFoundClusterActionError, StackNotFoundError):
raise NotFoundException(
f"Cluster '{cluster_name}' does not exist or belongs to an incompatible ParallelCluster major version."
)