pyiron_base.jobs.job.extension.server.generic.Server

class pyiron_base.jobs.job.extension.server.generic.Server(host: str | None = None, queue: str | None = None, cores: int = 1, threads: int = 1, gpus: int | None = None, run_mode: str = 'modal', new_hdf: bool = True, accept_crash: bool = False, run_time: int | None = None, memory_limit: str | None = None, qid: int | None = None, additional_arguments: dict = {}, conda_environment_name: str | None = None, conda_environment_path: str | None = None)

Bases: Lockable, HasDict

Generic Server object to handle the execution environment for the job

Parameters:

• host (str) – hostname of the local machine

• queue (str) – queue name of the currently selected queue

• cores (int) – number of cores

• run_mode (pyiron_base.server.runmode.Runmode) – mode of the job [‘modal’, ‘non_modal’, ‘queue’, ‘manual’]

• new_hdf (bool) – create a new HDF5 file [True/False] - default=True

• accept_crash (bool) – ignore execution errors raised by external executables - default False

• run_time (int) – run time limit in seconds for the job to finish - required for HPC job schedulers

• memory_limit (str) – memory required

• qid (int) – Queuing system ID - ID received from the HPC job scheduler

• additional_arguments (dict) – Additional arguments for the HPC job scheduler

• conda_environment_name (str) – Name of the conda environment

• conda_environment_path (str) – Path to the conda environment

.. attribute:: host

the hostname of the current system.

.. attribute:: queue

the queue selected for a current simulation.

.. attribute:: cores

the number of cores selected for the current simulation.

.. attribute:: run_time

the run time in seconds selected for the current simulation.

.. attribute:: run_mode

the run mode of the job [‘modal’, ‘non_modal’, ‘queue’, ‘manual’]

.. attribute:: memory_limit

the maximum amount of RAM allocated for the calculation in GB

.. attribute:: new_hdf

defines whether a subjob should be stored in the same HDF5 file or in a new one.

.. attribute:: executor

the executor can be used to execute the job object

.. attribute:: future

the concurrent.futures.Future object for monitoring the execution of the job object

__init__(host: str | None = None, queue: str | None = None, cores: int = 1, threads: int = 1, gpus: int | None = None, run_mode: str = 'modal', new_hdf: bool = True, accept_crash: bool = False, run_time: int | None = None, memory_limit: str | None = None, qid: int | None = None, additional_arguments: dict = {}, conda_environment_name: str | None = None, conda_environment_path: str | None = None)

Methods

__init__([host, queue, cores, threads, ...])
db_entry()	connect all the info regarding the server into a single word that can be used e.g. as entry in a database.
from_dict(obj_dict[, version])	Populate the object from the serialized object.
from_hdf(hdf[, group_name])	Recover Server object in HDF5 file :param hdf: HDF5 object :param group_name: node name in the HDF5 file
instantiate(obj_dict[, version])	Create a blank instance of this class.
list_queues()	List the available Job scheduler provided by the system.
lock([method])	Set read_only.
to_dict()	Reduce the object to a dictionary.
to_hdf(hdf[, group_name])	Store Server object in HDF5 file :param hdf: HDF5 object :param group_name: node name in the HDF5 file :type group_name: str
unlocked()	Unlock the object temporarily.
view_queues()	List the available Job scheduler provided by the system.

Attributes

accept_crash
additional_arguments
conda_environment_name	Get name of the conda environment the job should be executed in.
conda_environment_path	Get path of the conda environment the job should be executed in.
cores	The number of cores selected for the current simulation
executor	Executor to execute the job object this server object is attached to in the background.
future	Python concurrent.futures.Future object to track the status of the execution of the job this server object is attached to.
gpus	Total number of GPUs to use for this calculation.
memory_limit
new_hdf	New_hdf5 defines whether a subjob should be stored in the same HDF5 file or in a new one.
queue	The que selected for a current simulation
queue_id	Get the queue ID - the ID in the queuing system is most likely not the same as the job ID.
queue_list	List the available Job scheduler provided by the system.
queue_view	List the available Job scheduler provided by the system.
read_only	False if the object can currently be written to
run_mode	Get the run mode of the job
run_time	The run time in seconds selected for the current simulation
threads	The number of threads selected for the current simulation

property conda_environment_name: str

Get name of the conda environment the job should be executed in.

Returns:

name of the conda environment

Return type:

str

property conda_environment_path: str

Get path of the conda environment the job should be executed in.

Returns:

path of the conda environment

Return type:

str

property cores: int

The number of cores selected for the current simulation

Returns:

number of cores

Return type:

(int)

db_entry() → str

connect all the info regarding the server into a single word that can be used e.g. as entry in a database

Returns:

server info as single word

Return type:

(str)

property executor: Executor | None

Executor to execute the job object this server object is attached to in the background.

Return type:

concurrent.futures.Executor

from_dict(obj_dict: dict, version: str = None)

Populate the object from the serialized object.

Parameters:

• obj_dict (dict) – data previously returned from to_dict()

• version (str) – version tag written together with the data

from_hdf(hdf: pyiron_base.storage.hdfio.ProjectHDFio, group_name: str | None = None) → None

Recover Server object in HDF5 file :param hdf: HDF5 object :param group_name: node name in the HDF5 file

property future: Future | None

Python concurrent.futures.Future object to track the status of the execution of the job this server object is attached to. This is an internal pyiron feature and most users never have to interact with the future object directly.

Returns:

future object to track the status of the execution

Return type:

concurrent.futures.Future

property gpus: int | None

Total number of GPUs to use for this calculation.

Returns:

Total number of GPUs to use for this calculation.

Return type:

int

classmethod instantiate(obj_dict: dict, version: str = None) → Self

Create a blank instance of this class.

This can be used when some values are already necessary for the objects __init__.

Parameters:

• obj_dict (dict) – data previously returned from to_dict()

• version (str) – version tag written together with the data

Returns:

a blank instance of the object that is sufficiently initialized to call _from_dict() on it

Return type:

object

static list_queues() → list

List the available Job scheduler provided by the system.

Returns:

(list)

lock(method: Literal['error', 'warning'] | None = None)

Set read_only.

Objects may be safely locked multiple times without further effect.

Parameters:

method (str, either "error" or "warning") – if “error” raise an Locked exception if modification is attempted; if “warning” raise a LockedWarning warning; default is “error” or the value passed to the constructor.

Raises:

ValueError – if method is not an allowed value

property new_hdf: bool

New_hdf5 defines whether a subjob should be stored in the same HDF5 file or in a new one.

Returns:

[True / False]

Return type:

(bool)

property queue: str | None

The que selected for a current simulation

Returns:

schedulers_name

Return type:

(str)

property queue_id: int

Get the queue ID - the ID in the queuing system is most likely not the same as the job ID.

Returns:

queue ID

Return type:

int

property queue_list: list

List the available Job scheduler provided by the system.

Returns:

(list)

property queue_view: DataFrame

List the available Job scheduler provided by the system.

Returns:

(pandas.DataFrame)

property read_only: bool

False if the object can currently be written to

Setting this value will trigger _on_lock() and _on_unlock() if it changes.

Type:

bool

property run_mode: str

Get the run mode of the job

Returns:

[‘modal’, ‘non_modal’, ‘queue’, ‘manual’, ‘thread’, ‘worker’,

’interactive’, ‘interactive_non_modal’, ‘srun’, ‘executor’]

Return type:

(str/pyiron_base.server.runmode.Runmode)

property run_time: int

The run time in seconds selected for the current simulation

Returns:

run time in seconds

Return type:

(int)

property threads: int

The number of threads selected for the current simulation

Returns:

number of threads

Return type:

(int)

to_dict() → dict

Reduce the object to a dictionary.

Returns:

serialized state of this object

Return type:

dict

to_hdf(hdf: pyiron_base.storage.hdfio.ProjectHDFio, group_name: str | None = None) → None

Store Server object in HDF5 file :param hdf: HDF5 object :param group_name: node name in the HDF5 file :type group_name: str

unlocked() → _UnlockContext

Unlock the object temporarily.

Context manager returns this object again and relocks it after the with statement finished.

Note

lock() vs. unlocked()

There is a small asymmetry between these two methods. lock() can only be done once (meaningfully), while unlocked() is a context manager and can be called multiple times.

static view_queues() → DataFrame

List the available Job scheduler provided by the system.

Returns:

(pandas.DataFrame)