SchedulerOptions#
- class pyedb.workflows.job_manager.backend.job_submission.SchedulerOptions(/, **data: Any)#
Bases:
pydantic.BaseModelResource requirements and scheduler-specific directives.
All attributes are validated by
validate(), which is automatically called after instantiation.- Parameters:
- queue
str,optional Partition or queue name. Defaults to
"default".- time
str,optional Wall-time limit in
HH:MM:SSorD.HH:MM:SS. Defaults to"24:00:00".- nodes
int,optional Number of compute nodes. Defaults to
1.- tasks_per_node
int,optional Processes per node. Defaults to
1.- memory
str,optional Memory per node, e.g.
"4GB". Defaults to"4GB".- account
str,optional Account / project to charge.
- reservation
str,optional Advance reservation name.
- qos
str,optional Quality-of-service level.
- constraints
str,optional Node features, e.g.
"gpu".- exclusivebool,
optional Request whole nodes. Defaults to
False.- gpus
int,optional Number of GPUs. Defaults to
0.- gpu_type
str,optional GPU model, e.g.
"a100".- priority
str,optional Job priority:
Low,BelowNormal,Normal,AboveNormal,High. Defaults to"Normal".- email_notification
str,optional Address for status mails.
- run_as_administratorbool,
optional Elevated privileges (Windows HPC only). Defaults to
False.
- queue
- Raises:
ValueErrorOn any validation failure.
Overview#
Validate all scheduler options for correctness and consistency. |
Import detail#
from pyedb.workflows.job_manager.backend.job_submission import SchedulerOptions
Attribute detail#
Method detail#
- SchedulerOptions.validate_fields() None#
Validate all scheduler options for correctness and consistency.
- Raises:
ValueErrorIf any parameter is invalid or out of range.