Slurm Container

This workload (test_template_name is SlurmContainer) executes containerized workloads using Slurm with custom container configurations.

Test TOML example:

            

            
name = "my_container_test"
description = "Example Slurm container test"
test_template_name = "SlurmContainer"

[cmd_args]
image_path = "/path/to/container.sqsh"
cmd = "python train.py"
        

Test Scenario example:

            

            
name = "slurm-container-test"

[[Tests]]
id = "container.1"
num_nodes = 2
time_limit = "01:00:00"

test_name = "my_container_test"
        

Test-in-Scenario example:

            

            
name = "slurm-container-test"

[[Tests]]
id = "container.1"
num_nodes = 2
time_limit = "01:00:00"

name = "my_container_test"
description = "Example Slurm container test"
test_template_name = "SlurmContainer"

  [Tests.cmd_args]
  image_path = "/path/to/container.sqsh"
  cmd = "python train.py"
        

Command Arguments

class cloudai.workloads.slurm_container.slurm_container.SlurmContainerCmdArgs(*, docker_image_url: str, cmd: str, **extra_data: Any)[source]

Bases: CmdArgs

Command line arguments for a generic Slurm container test.

Test Definition

class cloudai.workloads.slurm_container.slurm_container.SlurmContainerTestDefinition(*, name: str, description: str, test_template_name: str, cmd_args: ~cloudai.workloads.slurm_container.slurm_container.SlurmContainerCmdArgs, extra_env_vars: dict[str, str | ~typing.List[str]] = {}, extra_cmd_args: dict[str, str] = {}, extra_container_mounts: list[str] = [], git_repos: list[~cloudai._core.installables.GitRepo] = [], nsys: ~cloudai.models.workload.NsysConfiguration | None = None, predictor: ~cloudai.models.workload.PredictorConfig | None = None, agent: str = 'grid_search', agent_steps: int = 1, agent_metrics: list[str] = ['default'], agent_reward_function: str = 'inverse', agent_config: dict[str, ~typing.Any] | None = None, extra_srun_args: list[str] = <factory>, scripts: list[~cloudai._core.installables.File] = <factory>)[source]

Bases: TestDefinition

Test definition for a generic Slurm container test.