#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
from typing import overload, Optional, Dict
from py4j.java_gateway import JavaObject, JVMView
from pyspark.util import _parse_memory
[docs]class ExecutorResourceRequest:
"""
An Executor resource request. This is used in conjunction with the ResourceProfile to
programmatically specify the resources needed for an RDD that will be applied at the
stage level.
This is used to specify what the resource requirements are for an Executor and how
Spark can find out specific details about those resources. Not all the parameters are
required for every resource type. Resources like GPUs are supported and have same limitations
as using the global spark configs spark.executor.resource.gpu.*. The amount, discoveryScript,
and vendor parameters for resources are all the same parameters a user would specify through the
configs: spark.executor.resource.{resourceName}.{amount, discoveryScript, vendor}.
For instance, a user wants to allocate an Executor with GPU resources on YARN. The user has
to specify the resource name (gpu), the amount or number of GPUs per Executor,
the discovery script would be specified so that when the Executor starts up it can
discovery what GPU addresses are available for it to use because YARN doesn't tell
Spark that, then vendor would not be used because its specific for Kubernetes.
See the configuration and cluster specific docs for more details.
Use :class:`pyspark.ExecutorResourceRequests` class as a convenience API.
.. versionadded:: 3.1.0
Parameters
----------
resourceName : str
Name of the resource
amount : str
Amount requesting
discoveryScript : str, optional
Optional script used to discover the resources. This is required on some
cluster managers that don't tell Spark the addresses of the resources
allocated. The script runs on Executors startup to discover the addresses
of the resources available.
vendor : str, optional
Vendor, required for some cluster managers
See Also
--------
:class:`pyspark.resource.ResourceProfile`
Notes
-----
This API is evolving.
"""
def __init__(
self,
resourceName: str,
amount: int,
discoveryScript: str = "",
vendor: str = "",
):
self._name = resourceName
self._amount = amount
self._discovery_script = discoveryScript
self._vendor = vendor
@property
def resourceName(self) -> str:
"""
Returns
-------
str
Name of the resource
"""
return self._name
@property
def amount(self) -> int:
"""
Returns
-------
str
Amount requesting
"""
return self._amount
@property
def discoveryScript(self) -> str:
"""
Returns
-------
str
Amount requesting
"""
return self._discovery_script
@property
def vendor(self) -> str:
"""
Returns
-------
str
Vendor, required for some cluster managers
"""
return self._vendor
[docs]class ExecutorResourceRequests:
"""
A set of Executor resource requests. This is used in conjunction with the
:class:`pyspark.resource.ResourceProfileBuilder` to programmatically specify the
resources needed for an RDD that will be applied at the stage level.
.. versionadded:: 3.1.0
See Also
--------
:class:`pyspark.resource.ResourceProfile`
Notes
-----
This API is evolving.
"""
_CORES = "cores"
_MEMORY = "memory"
_OVERHEAD_MEM = "memoryOverhead"
_PYSPARK_MEM = "pyspark.memory"
_OFFHEAP_MEM = "offHeap"
@overload
def __init__(self, _jvm: JVMView):
...
@overload
def __init__(
self,
_jvm: None = ...,
_requests: Optional[Dict[str, ExecutorResourceRequest]] = ...,
):
...
def __init__(
self,
_jvm: Optional[JVMView] = None,
_requests: Optional[Dict[str, ExecutorResourceRequest]] = None,
):
from pyspark import SparkContext
_jvm = _jvm or SparkContext._jvm
if _jvm is not None:
self._java_executor_resource_requests = (
_jvm.org.apache.spark.resource.ExecutorResourceRequests()
)
if _requests is not None:
for k, v in _requests.items():
if k == self._MEMORY:
self._java_executor_resource_requests.memory(str(v.amount))
elif k == self._OVERHEAD_MEM:
self._java_executor_resource_requests.memoryOverhead(str(v.amount))
elif k == self._PYSPARK_MEM:
self._java_executor_resource_requests.pysparkMemory(str(v.amount))
elif k == self._CORES:
self._java_executor_resource_requests.cores(v.amount)
else:
self._java_executor_resource_requests.resource(
v.resourceName, v.amount, v.discoveryScript, v.vendor
)
else:
self._java_executor_resource_requests = None
self._executor_resources: Dict[str, ExecutorResourceRequest] = {}
def memory(self, amount: str) -> "ExecutorResourceRequests":
"""
Specify heap memory. The value specified will be converted to MiB.
This is a convenient API to add :class:`ExecutorResourceRequest` for "memory" resource.
Parameters
----------
amount : str
Amount of memory. In the same format as JVM memory strings (e.g. 512m, 2g).
Default unit is MiB if not specified.
Returns
-------
:class:`ExecutorResourceRequests`
"""
if self._java_executor_resource_requests is not None:
self._java_executor_resource_requests.memory(amount)
else:
self._executor_resources[self._MEMORY] = ExecutorResourceRequest(
self._MEMORY, _parse_memory(amount)
)
return self
def memoryOverhead(self, amount: str) -> "ExecutorResourceRequests":
"""
Specify overhead memory. The value specified will be converted to MiB.
This is a convenient API to add :class:`ExecutorResourceRequest` for "memoryOverhead"
resource.
Parameters
----------
amount : str
Amount of memory. In the same format as JVM memory strings (e.g. 512m, 2g).
Default unit is MiB if not specified.
Returns
-------
:class:`ExecutorResourceRequests`
"""
if self._java_executor_resource_requests is not None:
self._java_executor_resource_requests.memoryOverhead(amount)
else:
self._executor_resources[self._OVERHEAD_MEM] = ExecutorResourceRequest(
self._OVERHEAD_MEM, _parse_memory(amount)
)
return self
def pysparkMemory(self, amount: str) -> "ExecutorResourceRequests":
"""
Specify pyspark memory. The value specified will be converted to MiB.
This is a convenient API to add :class:`ExecutorResourceRequest` for "pyspark.memory"
resource.
Parameters
----------
amount : str
Amount of memory. In the same format as JVM memory strings (e.g. 512m, 2g).
Default unit is MiB if not specified.
Returns
-------
:class:`ExecutorResourceRequests`
"""
if self._java_executor_resource_requests is not None:
self._java_executor_resource_requests.pysparkMemory(amount)
else:
self._executor_resources[self._PYSPARK_MEM] = ExecutorResourceRequest(
self._PYSPARK_MEM, _parse_memory(amount)
)
return self
def offheapMemory(self, amount: str) -> "ExecutorResourceRequests":
"""
Specify off heap memory. The value specified will be converted to MiB.
This value only take effect when MEMORY_OFFHEAP_ENABLED is true.
This is a convenient API to add :class:`ExecutorResourceRequest` for "offHeap"
resource.
Parameters
----------
amount : str
Amount of memory. In the same format as JVM memory strings (e.g. 512m, 2g).
Default unit is MiB if not specified.
Returns
-------
:class:`ExecutorResourceRequests`
"""
if self._java_executor_resource_requests is not None:
self._java_executor_resource_requests.offHeapMemory(amount)
else:
self._executor_resources[self._OFFHEAP_MEM] = ExecutorResourceRequest(
self._OFFHEAP_MEM, _parse_memory(amount)
)
return self
def cores(self, amount: int) -> "ExecutorResourceRequests":
"""
Specify number of cores per Executor.
This is a convenient API to add :class:`ExecutorResourceRequest` for "cores" resource.
Parameters
----------
amount : int
Number of cores to allocate per Executor.
Returns
-------
:class:`ExecutorResourceRequests`
"""
if self._java_executor_resource_requests is not None:
self._java_executor_resource_requests.cores(amount)
else:
self._executor_resources[self._CORES] = ExecutorResourceRequest(self._CORES, amount)
return self
def resource(
self,
resourceName: str,
amount: int,
discoveryScript: str = "",
vendor: str = "",
) -> "ExecutorResourceRequests":
"""
Amount of a particular custom resource(GPU, FPGA, etc) to use. The resource names supported
correspond to the regular Spark configs with the prefix removed. For instance, resources
like GPUs are gpu (spark configs `spark.executor.resource.gpu.*`). If you pass in a resource
that the cluster manager doesn't support the result is undefined, it may error or may just
be ignored.
This is a convenient API to add :class:`ExecutorResourceRequest` for custom resources.
Parameters
----------
resourceName : str
Name of the resource.
amount : str
amount of that resource per executor to use.
discoveryScript : str, optional
Optional script used to discover the resources. This is required on
some cluster managers that don't tell Spark the addresses of
the resources allocated. The script runs on Executors startup to
of the resources available.
vendor : str
Optional vendor, required for some cluster managers
Returns
-------
:class:`ExecutorResourceRequests`
"""
if self._java_executor_resource_requests is not None:
self._java_executor_resource_requests.resource(
resourceName, amount, discoveryScript, vendor
)
else:
self._executor_resources[resourceName] = ExecutorResourceRequest(
resourceName, amount, discoveryScript, vendor
)
return self
@property
def requests(self) -> Dict[str, ExecutorResourceRequest]:
"""
Returns
-------
dict
Returns all the resource requests for the executor.
"""
if self._java_executor_resource_requests is not None:
result = {}
execRes = self._java_executor_resource_requests.requestsJMap()
for k, v in execRes.items():
result[k] = ExecutorResourceRequest(
v.resourceName(), v.amount(), v.discoveryScript(), v.vendor()
)
return result
else:
return self._executor_resources
[docs]class TaskResourceRequest:
"""
A task resource request. This is used in conjunction with the
:class:`pyspark.resource.ResourceProfile` to programmatically specify the resources
needed for an RDD that will be applied at the stage level. The amount is specified
as a float to allow for saying you want more than 1 task per resource. Valid values
are less than or equal to 0.5 or whole numbers.
Use :class:`pyspark.resource.TaskResourceRequests` class as a convenience API.
Parameters
----------
resourceName : str
Name of the resource
amount : float
Amount requesting as a float to support fractional resource requests.
Valid values are less than or equal to 0.5 or whole numbers. This essentially
lets you configure X number of tasks to run on a single resource,
ie amount equals 0.5 translates into 2 tasks per resource address.
.. versionadded:: 3.1.0
See Also
--------
:class:`pyspark.resource.ResourceProfile`
Notes
-----
This API is evolving.
"""
def __init__(self, resourceName: str, amount: float):
self._name = resourceName
self._amount = float(amount)
@property
def resourceName(self) -> str:
"""
Returns
-------
str
Name of the resource.
"""
return self._name
@property
def amount(self) -> float:
"""
Returns
-------
str
Amount requesting as a float to support fractional resource requests.
"""
return self._amount
[docs]class TaskResourceRequests:
"""
A set of task resource requests. This is used in conjunction with the
:class:`pyspark.resource.ResourceProfileBuilder` to programmatically specify the resources
needed for an RDD that will be applied at the stage level.
.. versionadded:: 3.1.0
See Also
--------
:class:`pyspark.resource.ResourceProfile`
Notes
-----
This API is evolving.
"""
_CPUS = "cpus"
@overload
def __init__(self, _jvm: JVMView):
...
@overload
def __init__(
self,
_jvm: None = ...,
_requests: Optional[Dict[str, TaskResourceRequest]] = ...,
):
...
def __init__(
self,
_jvm: Optional[JVMView] = None,
_requests: Optional[Dict[str, TaskResourceRequest]] = None,
):
from pyspark import SparkContext
_jvm = _jvm or SparkContext._jvm
if _jvm is not None:
self._java_task_resource_requests: Optional[
JavaObject
] = _jvm.org.apache.spark.resource.TaskResourceRequests()
if _requests is not None:
for k, v in _requests.items():
if k == self._CPUS:
self._java_task_resource_requests.cpus(int(v.amount))
else:
self._java_task_resource_requests.resource(v.resourceName, v.amount)
else:
self._java_task_resource_requests = None
self._task_resources: Dict[str, TaskResourceRequest] = {}
def cpus(self, amount: int) -> "TaskResourceRequests":
"""
Specify number of cpus per Task.
This is a convenient API to add :class:`TaskResourceRequest` for cpus.
Parameters
----------
amount : int
Number of cpus to allocate per Task.
Returns
-------
:class:`TaskResourceRequests`
"""
if self._java_task_resource_requests is not None:
self._java_task_resource_requests.cpus(amount)
else:
self._task_resources[self._CPUS] = TaskResourceRequest(self._CPUS, amount)
return self
def resource(self, resourceName: str, amount: float) -> "TaskResourceRequests":
"""
Amount of a particular custom resource(GPU, FPGA, etc) to use.
This is a convenient API to add :class:`TaskResourceRequest` for custom resources.
Parameters
----------
resourceName : str
Name of the resource.
amount : float
Amount requesting as a float to support fractional resource requests.
Valid values are less than or equal to 0.5 or whole numbers. This essentially
lets you configure X number of tasks to run on a single resource,
ie amount equals 0.5 translates into 2 tasks per resource address.
Returns
-------
:class:`TaskResourceRequests`
"""
if self._java_task_resource_requests is not None:
self._java_task_resource_requests.resource(resourceName, float(amount))
else:
self._task_resources[resourceName] = TaskResourceRequest(resourceName, amount)
return self
@property
def requests(self) -> Dict[str, TaskResourceRequest]:
"""
Returns
-------
dict
Returns all the resource requests for the task.
"""
if self._java_task_resource_requests is not None:
result = {}
taskRes = self._java_task_resource_requests.requestsJMap()
for k, v in taskRes.items():
result[k] = TaskResourceRequest(v.resourceName(), v.amount())
return result
else:
return self._task_resources