Source code for aiida.schedulers.datastructures

# -*- coding: utf-8 -*-
###########################################################################
# Copyright (c), The AiiDA team. All rights reserved.                     #
# This file is part of the AiiDA code.                                    #
#                                                                         #
# The code is hosted on GitHub at https://github.com/aiidateam/aiida-core #
# For further information on the license, see the LICENSE.txt file        #
# For further information please visit http://www.aiida.net               #
###########################################################################
"""Data structures used by `Scheduler` instances.

In particular, there is the definition of possible job states (job_states),
the data structure to be filled for job submission (JobTemplate), and
the data structure that is returned when querying for jobs in the scheduler
(JobInfo).
"""
import abc
import enum

from aiida.common import AIIDA_LOGGER
from aiida.common.extendeddicts import AttributeDict, DefaultFieldsAttributeDict

SCHEDULER_LOGGER = AIIDA_LOGGER.getChild('scheduler')

__all__ = (
    'JobState', 'JobResource', 'JobTemplate', 'JobInfo', 'NodeNumberJobResource', 'ParEnvJobResource', 'MachineInfo'
)


[docs]class JobState(enum.Enum): """Enumeration of possible scheduler states of a CalcJob. There is no FAILED state as every completed job is put in DONE, regardless of success. """ UNDETERMINED = 'undetermined' QUEUED = 'queued' QUEUED_HELD = 'queued held' RUNNING = 'running' SUSPENDED = 'suspended' DONE = 'done'
[docs]class JobResource(DefaultFieldsAttributeDict, metaclass=abc.ABCMeta): """Data structure to store job resources. Each `Scheduler` implementation must define the `_job_resource_class` attribute to be a subclass of this class. It should at least define the `get_tot_num_mpiprocs` method, plus a constructor to accept its set of variables. Typical attributes are: * ``num_machines`` * ``num_mpiprocs_per_machine`` or (e.g. for SGE) * ``tot_num_mpiprocs`` * ``parallel_env`` The constructor should take care of checking the values. The init should raise only ValueError or TypeError on invalid parameters. """ _default_fields = tuple()
[docs] @abc.abstractclassmethod def validate_resources(cls, **kwargs): """Validate the resources against the job resource class of this scheduler. :param kwargs: dictionary of values to define the job resources :raises ValueError: if the resources are invalid or incomplete :return: optional tuple of parsed resource settings """
[docs] @classmethod def get_valid_keys(cls): """Return a list of valid keys to be passed to the constructor.""" return list(cls._default_fields)
[docs] @abc.abstractclassmethod def accepts_default_mpiprocs_per_machine(cls): """Return True if this subclass accepts a `default_mpiprocs_per_machine` key, False otherwise."""
[docs] @abc.abstractmethod def get_tot_num_mpiprocs(self): """Return the total number of cpus of this job resource."""
[docs]class NodeNumberJobResource(JobResource): """`JobResource` for schedulers that support the specification of a number of nodes and cpus per node.""" _default_fields = ( 'num_machines', 'num_mpiprocs_per_machine', 'num_cores_per_machine', 'num_cores_per_mpiproc', )
[docs] @classmethod def validate_resources(cls, **kwargs): """Validate the resources against the job resource class of this scheduler. :param kwargs: dictionary of values to define the job resources :return: attribute dictionary with the parsed parameters populated :raises ValueError: if the resources are invalid or incomplete """ resources = AttributeDict() def is_greater_equal_one(parameter): value = getattr(resources, parameter, None) if value is not None and value < 1: raise ValueError(f'`{parameter}` must be greater than or equal to one.') # Validate that all fields are valid integers if they are specified, otherwise initialize them to `None` for parameter in list(cls._default_fields) + ['tot_num_mpiprocs']: value = kwargs.pop(parameter, None) if value is None: setattr(resources, parameter, None) else: try: setattr(resources, parameter, int(value)) except ValueError: raise ValueError(f'`{parameter}` must be an integer when specified') if kwargs: raise ValueError(f"these parameters were not recognized: {', '.join(list(kwargs.keys()))}") # At least two of the following parameters need to be defined as non-zero if [resources.num_machines, resources.num_mpiprocs_per_machine, resources.tot_num_mpiprocs].count(None) > 1: raise ValueError( 'At least two among `num_machines`, `num_mpiprocs_per_machine` or `tot_num_mpiprocs` must be specified.' ) for parameter in ['num_machines', 'num_mpiprocs_per_machine']: is_greater_equal_one(parameter) # Here we now that at least two of the three required variables are defined and greater equal than one. if resources.num_machines is None: resources.num_machines = resources.tot_num_mpiprocs // resources.num_mpiprocs_per_machine elif resources.num_mpiprocs_per_machine is None: resources.num_mpiprocs_per_machine = resources.tot_num_mpiprocs // resources.num_machines elif resources.tot_num_mpiprocs is None: resources.tot_num_mpiprocs = resources.num_mpiprocs_per_machine * resources.num_machines if resources.tot_num_mpiprocs != resources.num_mpiprocs_per_machine * resources.num_machines: raise ValueError('`tot_num_mpiprocs` is not equal to `num_mpiprocs_per_machine * num_machines`.') is_greater_equal_one('num_mpiprocs_per_machine') is_greater_equal_one('num_machines') return resources
[docs] def __init__(self, **kwargs): """Initialize the job resources from the passed arguments. :raises ValueError: if the resources are invalid or incomplete """ resources = self.validate_resources(**kwargs) super().__init__(resources)
[docs] @classmethod def get_valid_keys(cls): """Return a list of valid keys to be passed to the constructor.""" return super().get_valid_keys() + ['tot_num_mpiprocs']
[docs] @classmethod def accepts_default_mpiprocs_per_machine(cls): """Return True if this subclass accepts a `default_mpiprocs_per_machine` key, False otherwise.""" return True
[docs] def get_tot_num_mpiprocs(self): """Return the total number of cpus of this job resource.""" return self.num_machines * self.num_mpiprocs_per_machine
[docs]class ParEnvJobResource(JobResource): """`JobResource` for schedulers that support the specification of a parallel environment and number of MPI procs.""" _default_fields = ( 'parallel_env', 'tot_num_mpiprocs', )
[docs] @classmethod def validate_resources(cls, **kwargs): """Validate the resources against the job resource class of this scheduler. :param kwargs: dictionary of values to define the job resources :return: attribute dictionary with the parsed parameters populated :raises ValueError: if the resources are invalid or incomplete """ resources = AttributeDict() try: resources.parallel_env = kwargs.pop('parallel_env') except KeyError: raise ValueError('`parallel_env` must be specified and must be a string') else: if not isinstance(resources.parallel_env, str): raise ValueError('`parallel_env` must be specified and must be a string') try: resources.tot_num_mpiprocs = int(kwargs.pop('tot_num_mpiprocs')) except (KeyError, TypeError, ValueError): raise ValueError('`tot_num_mpiprocs` must be specified and must be an integer') if resources.tot_num_mpiprocs < 1: raise ValueError('`tot_num_mpiprocs` must be greater than or equal to one.') if kwargs: raise ValueError(f"these parameters were not recognized: {', '.join(list(kwargs.keys()))}") return resources
[docs] def __init__(self, **kwargs): """ Initialize the job resources from the passed arguments (the valid keys can be obtained with the function self.get_valid_keys()). :raises ValueError: if the resources are invalid or incomplete """ resources = self.validate_resources(**kwargs) super().__init__(resources)
[docs] @classmethod def accepts_default_mpiprocs_per_machine(cls): """Return True if this subclass accepts a `default_mpiprocs_per_machine` key, False otherwise.""" return False
[docs] def get_tot_num_mpiprocs(self): """Return the total number of cpus of this job resource.""" return self.tot_num_mpiprocs
[docs]class JobTemplate(DefaultFieldsAttributeDict): # pylint: disable=too-many-instance-attributes """A template for submitting jobs to a scheduler. This contains all required information to create the job header. The required fields are: working_directory, job_name, num_machines, num_mpiprocs_per_machine, argv. Fields: * ``shebang line``: The first line of the submission script * ``submit_as_hold``: if set, the job will be in a 'hold' status right after the submission * ``rerunnable``: if the job is rerunnable (boolean) * ``job_environment``: a dictionary with environment variables to set before the execution of the code. * ``working_directory``: the working directory for this job. During submission, the transport will first do a 'chdir' to this directory, and then possibly set a scheduler parameter, if this is supported by the scheduler. * ``email``: an email address for sending emails on job events. * ``email_on_started``: if True, ask the scheduler to send an email when the job starts. * ``email_on_terminated``: if True, ask the scheduler to send an email when the job ends. This should also send emails on job failure, when possible. * ``job_name``: the name of this job. The actual name of the job can be different from the one specified here, e.g. if there are unsupported characters, or the name is too long. * ``sched_output_path``: a (relative) file name for the stdout of this job * ``sched_error_path``: a (relative) file name for the stdout of this job * ``sched_join_files``: if True, write both stdout and stderr on the same file (the one specified for stdout) * ``queue_name``: the name of the scheduler queue (sometimes also called partition), on which the job will be submitted. * ``account``: the name of the scheduler account (sometimes also called projectid), on which the job will be submitted. * ``qos``: the quality of service of the scheduler account, on which the job will be submitted. * ``job_resource``: a suitable :py:class:`JobResource` subclass with information on how many nodes and cpus it should use. It must be an instance of the ``aiida.schedulers.Scheduler.job_resource_class`` class. Use the Scheduler.create_job_resource method to create it. * ``num_machines``: how many machines (or nodes) should be used * ``num_mpiprocs_per_machine``: how many MPI procs should be used on each machine (or node). * ``priority``: a priority for this job. Should be in the format accepted by the specific scheduler. * ``max_memory_kb``: The maximum amount of memory the job is allowed to allocate ON EACH NODE, in kilobytes * ``max_wallclock_seconds``: The maximum wall clock time that all processes of a job are allowed to exist, in seconds * ``custom_scheduler_commands``: a string that will be inserted right after the last scheduler command, and before any other non-scheduler command; useful if some specific flag needs to be added and is not supported by the plugin * ``prepend_text``: a (possibly multi-line) string to be inserted in the scheduler script before the main execution line * ``append_text``: a (possibly multi-line) string to be inserted in the scheduler script after the main execution line * ``import_sys_environment``: import the system environment variables * ``codes_info``: a list of aiida.common.datastructures.CalcInfo objects. Each contains the information necessary to run a single code. At the moment, it can contain: * ``cmdline_parameters``: a list of strings with the command line arguments of the program to run. This is the main program to be executed. NOTE: The first one is the executable name. For MPI runs, this will probably be "mpirun" or a similar program; this has to be chosen at a upper level. * ``stdin_name``: the (relative) file name to be used as stdin for the program specified with argv. * ``stdout_name``: the (relative) file name to be used as stdout for the program specified with argv. * ``stderr_name``: the (relative) file name to be used as stderr for the program specified with argv. * ``join_files``: if True, stderr is redirected on the same file specified for stdout. * ``codes_run_mode``: sets the run_mode with which the (multiple) codes have to be executed. For example, parallel execution:: mpirun -np 8 a.x & mpirun -np 8 b.x & wait The serial execution would be without the &'s. Values are given by aiida.common.datastructures.CodeRunMode. """ _default_fields = ( 'shebang', 'submit_as_hold', 'rerunnable', 'job_environment', 'working_directory', 'email', 'email_on_started', 'email_on_terminated', 'job_name', 'sched_output_path', 'sched_error_path', 'sched_join_files', 'queue_name', 'account', 'qos', 'job_resource', 'priority', 'max_memory_kb', 'max_wallclock_seconds', 'custom_scheduler_commands', 'prepend_text', 'append_text', 'import_sys_environment', 'codes_run_mode', 'codes_info', )
[docs]class MachineInfo(DefaultFieldsAttributeDict): """ Similarly to what is defined in the DRMAA v.2 as SlotInfo; this identifies each machine (also called 'node' on some schedulers) on which a job is running, and how many CPUs are being used. (Some of them could be undefined) * ``name``: name of the machine * ``num_cpus``: number of cores used by the job on this machine * ``num_mpiprocs``: number of MPI processes used by the job on this machine """ _default_fields = ( 'name', 'num_mpiprocs', 'num_cpus', )
[docs]class JobInfo(DefaultFieldsAttributeDict): # pylint: disable=too-many-instance-attributes """ Contains properties for a job in the queue. Most of the fields are taken from DRMAA v.2. Note that default fields may be undefined. This is an expected behavior and the application must cope with this case. An example for instance is the exit_status for jobs that have not finished yet; or features not supported by the given scheduler. Fields: * ``job_id``: the job ID on the scheduler * ``title``: the job title, as known by the scheduler * ``exit_status``: the exit status of the job as reported by the operating system on the execution host * ``terminating_signal``: the UNIX signal that was responsible for the end of the job. * ``annotation``: human-readable description of the reason for the job being in the current state or substate. * ``job_state``: the job state (one of those defined in ``aiida.schedulers.datastructures.JobState``) * ``job_substate``: a string with the implementation-specific sub-state * ``allocated_machines``: a list of machines used for the current job. This is a list of :py:class:`aiida.schedulers.datastructures.MachineInfo` objects. * ``job_owner``: the job owner as reported by the scheduler * ``num_mpiprocs``: the *total* number of requested MPI procs * ``num_cpus``: the *total* number of requested CPUs (cores) [may be undefined] * ``num_machines``: the number of machines (i.e., nodes), required by the job. If ``allocated_machines`` is not None, this number must be equal to ``len(allocated_machines)``. Otherwise, for schedulers not supporting the retrieval of the full list of allocated machines, this attribute can be used to know at least the number of machines. * ``queue_name``: The name of the queue in which the job is queued or running. * ``account``: The account/projectid in which the job is queued or running in. * ``qos``: The quality of service in which the job is queued or running in. * ``wallclock_time_seconds``: the accumulated wallclock time, in seconds * ``requested_wallclock_time_seconds``: the requested wallclock time, in seconds * ``cpu_time``: the accumulated cpu time, in seconds * ``submission_time``: the absolute time at which the job was submitted, of type datetime.datetime * ``dispatch_time``: the absolute time at which the job first entered the 'started' state, of type datetime.datetime * ``finish_time``: the absolute time at which the job first entered the 'finished' state, of type datetime.datetime """ _default_fields = ( 'job_id', 'title', 'exit_status', 'terminating_signal', 'annotation', 'job_state', 'job_substate', 'allocated_machines', 'job_owner', 'num_mpiprocs', 'num_cpus', 'num_machines', 'queue_name', 'account', 'qos', 'wallclock_time_seconds', 'requested_wallclock_time_seconds', 'cpu_time', 'submission_time', 'dispatch_time', 'finish_time' ) # If some fields require special serializers, specify them here. # You then need to define also the respective _serialize_FIELDTYPE and # _deserialize_FIELDTYPE methods _special_serializers = { 'submission_time': 'date', 'dispatch_time': 'date', 'finish_time': 'date', 'job_state': 'job_state', }
[docs] @staticmethod def _serialize_job_state(job_state): """Return the serialized value of the JobState instance.""" if not isinstance(job_state, JobState): raise TypeError(f'invalid type for value {job_state}, should be an instance of `JobState`') return job_state.value
[docs] @staticmethod def _deserialize_job_state(job_state): """Return an instance of JobState from the job_state string.""" return JobState(job_state)
[docs] @staticmethod def _serialize_date(value): """ Serialise a data value :param value: The value to serialise :return: The serialised value """ import datetime import pytz if value is None: return value if not isinstance(value, datetime.datetime): raise TypeError('Invalid type for the date, should be a datetime') # is_naive check from django.utils.timezone if value.tzinfo is None or value.tzinfo.utcoffset(value) is None: SCHEDULER_LOGGER.debug('Datetime to serialize in JobInfo is naive, this should be fixed!') # v = v.replace(tzinfo = pytz.utc) return {'date': value.strftime('%Y-%m-%dT%H:%M:%S.%f'), 'timezone': None} return {'date': value.astimezone(pytz.utc).strftime('%Y-%m-%dT%H:%M:%S.%f'), 'timezone': 'UTC'}
[docs] @staticmethod def _deserialize_date(value): """ Deserialise a date :param value: The date vlue :return: The deserialised date """ import datetime import pytz if value is None: return value if value['timezone'] is None: # naive date return datetime.datetime.strptime(value['date'], '%Y-%m-%dT%H:%M:%S.%f') if value['timezone'] == 'UTC': return datetime.datetime.strptime(value['date'], '%Y-%m-%dT%H:%M:%S.%f').replace(tzinfo=pytz.utc) # Try your best return datetime.datetime.strptime(value['date'], '%Y-%m-%dT%H:%M:%S.%f').replace(tzinfo=pytz.timezone(value['timezone']))
[docs] @classmethod def serialize_field(cls, value, field_type): """ Serialise a particular field value :param value: The value to serialise :param field_type: The field type :return: The serialised value """ if field_type is None: return value serializer_method = getattr(cls, f'_serialize_{field_type}') return serializer_method(value)
[docs] @classmethod def deserialize_field(cls, value, field_type): """ Deserialise the value of a particular field with a type :param value: The value :param field_type: The field type :return: The deserialised value """ if field_type is None: return value deserializer_method = getattr(cls, f'_deserialize_{field_type}') return deserializer_method(value)
[docs] def serialize(self): """ Serialize the current data (as obtained by ``self.get_dict()``) into a JSON string. :return: A string with serialised representation of the current data. """ from aiida.common import json return json.dumps(self.get_dict())
[docs] def get_dict(self): """ Serialise the current data into a dictionary that is JSON-serializable. :return: A dictionary """ return {k: self.serialize_field(v, self._special_serializers.get(k, None)) for k, v in self.items()}
[docs] @classmethod def load_from_dict(cls, data): """ Create a new instance loading the values from serialised data in dictionary form :param data: The dictionary with the data to load from """ instance = cls() for key, value in data.items(): instance[key] = cls.deserialize_field(value, cls._special_serializers.get(key, None)) return instance
[docs] @classmethod def load_from_serialized(cls, data): """ Create a new instance loading the values from JSON-serialised data as a string :param data: The string with the JSON-serialised data to load from """ from aiida.common import json return cls.load_from_dict(json.loads(data))