# Copyright 2020 John Harwell, All rights reserved.
#
# SPDX-License-Identifier: MIT
"""Common cmdline classes for the various HPC plugins.
"""
# Core packages
import argparse
import typing as tp
# 3rd party packages
# Project packages
from sierra.core import types, cmdline
[docs]class HPCCmdline(cmdline.BaseCmdline):
[docs] def __init__(self, stages: tp.List[int]) -> None:
self.parser = argparse.ArgumentParser(add_help=False,
allow_abbrev=False)
self.scaffold_cli()
self.init_cli(stages)
[docs] def scaffold_cli(self) -> None:
desc = ("For platforms which are simulators (and can "
"therefore be run in HPC environments).")
self.hpc = self.parser.add_argument_group('HPC options', desc)
[docs] def init_cli(self, stages: tp.List[int]) -> None:
if 2 in stages:
self.init_stage2()
[docs] def init_stage2(self) -> None:
"""Add HPC cmdline options.
Options may be interpreted differently between :term:`Platforms
<Platform>`, or ignored, depending. These include:
- ``--exec-jobs-per-node``
- ``--exec-no-devnull``
- ``--exec-resume``
- ``--exec-strict``
"""
self.hpc.add_argument("--exec-jobs-per-node",
help="""
Specify the maximum number of parallel jobs to run
per allocated node. By default this is computed
from the selected HPC environment for maximum
throughput given the desired ``--n-runs`` and CPUs
per allocated node. However, for some environments
being able to override the computed default can be
useful.
""" + self.stage_usage_doc([2]),
type=int,
default=None)
self.hpc.add_argument("--exec-devnull",
help="""
Redirect ALL output from simulations to
/dev/null. Useful for platform where you can't
disable all INFO messages at compile time, and
don't want to have to grep through lots of
redundant stdout files to see if there were any
errors.
""" + self.stage_usage_doc([1, 2]),
action='store_true',
dest='exec_devnull',
default=True)
self.hpc.add_argument("--exec-no-devnull",
help="""
Don't redirect ALL output from simulations to
/dev/null. Useful for platform where you can't
disable all INFO messages at compile time, and
don't want to have to grep through lots of
redundant stdout files to see if there were any
errors.
""" + self.stage_usage_doc([1, 2]),
action='store_false',
dest='exec_devnull')
self.hpc.add_argument("--exec-resume",
help="""
Resume a batch experiment that was
killed/stopped/etc last time SIERRA was run.
""" + self.stage_usage_doc([2]),
action='store_true',
default=False)
self.hpc.add_argument("--exec-strict",
help="""
If passed, then if any experimental commands fail
during stage 2 SIERRA will exit, rather than try
to keep going and execute the rest of the
experiments.
Useful for:
- "Correctness by construction" experiments, where
you know if SIERRA doesn't crash and it makes it
to the end of your batch experiment then none of
the individual experiments crashed.
- CI pipelines
""",
action='store_true')
[docs] @staticmethod
def cmdopts_update(cli_args: argparse.Namespace,
cmdopts: types.Cmdopts) -> None:
"""Update cmdopts dictionary with the HPC-specific cmdline options.
"""
updates = {
# Multistage
'exec_devnull': cli_args.exec_devnull,
'exec_jobs_per_node': cli_args.exec_jobs_per_node,
'exec_resume': cli_args.exec_resume,
'exec_strict': cli_args.exec_strict
}
cmdopts.update(updates)
__api__ = [
'HPCCmdline',
]