in src/slurm_plugin/resume.py [0:0]
def main():
default_log_file = "/var/log/parallelcluster/slurm_resume.log"
logging.basicConfig(
filename=default_log_file,
level=logging.INFO,
format="%(asctime)s - %(process)d - [%(name)s:%(funcName)s] - %(levelname)s - %(message)s",
)
log.info("ResumeProgram startup.")
parser = argparse.ArgumentParser()
parser.add_argument("nodes", help="Nodes to burst")
args = parser.parse_args()
try:
config_file = os.environ.get("CONFIG_FILE", os.path.join(CONFIG_FILE_DIR, "parallelcluster_slurm_resume.conf"))
resume_config = SlurmResumeConfig(config_file)
try:
# Configure root logger
fileConfig(resume_config.logging_config, disable_existing_loggers=False)
except Exception as e:
log.warning(
"Unable to configure logging from %s, using default settings and writing to %s.\nException: %s",
resume_config.logging_config,
default_log_file,
e,
)
log.info("ResumeProgram config: %s", resume_config)
_resume(args.nodes, resume_config, _get_slurm_resume())
log.info("ResumeProgram finished.")
except Exception as e:
log.exception("Encountered exception when requesting instances for %s: %s", args.nodes, e)
_handle_failed_nodes(args.nodes)