“sinfo and a submitted GPU job both expose the approved GPU count” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is container bind mounts and permitted storage paths are misconfigured. The option “submitted container job reads and writes the expected shared paths” is more directly tied to the required outcome.

“an sbatch container job starts successfully with the intended image” addresses slurm SPANK plugin integration for Pyxis is incomplete, but the current hardware or infrastructure issue is container bind mounts and permitted storage paths are misconfigured. The option “submitted container job reads and writes the expected shared paths” is more directly tied to the required outcome.

“nvidia-smi inside the scheduled container shows only the assigned GPU devices” addresses container GPU exposure is not integrated with the scheduler allocation, but the current hardware or infrastructure issue is container bind mounts and permitted storage paths are misconfigured. The option “submitted container job reads and writes the expected shared paths” is more directly tied to the required outcome.

“configure consistent GRES and scheduling constraints for GPU requests” addresses slurm partition and GPU resource constraints were not enforced, but the current hardware or infrastructure issue is scheduler queues and limits do not represent workload isolation policy. The option “set appropriate partitions reservations and fair-share limits for competing jobs” is more directly tied to the required outcome.

“apply supported cgroup or device isolation in Slurm and its container integration” addresses gPU device access is not constrained to the scheduler allocation, but the current hardware or infrastructure issue is scheduler queues and limits do not represent workload isolation policy. The option “set appropriate partitions reservations and fair-share limits for competing jobs” is more directly tied to the required outcome.

Useful for container bind mounts and permitted storage paths are misconfigured, but it does not remove the specific condition: scheduler queues and limits do not represent workload isolation policy.

“a process cannot access a GPU not granted to its job allocation” addresses gPU device access is not constrained to the scheduler allocation, but the current hardware or infrastructure issue is scheduler queues and limits do not represent workload isolation policy. The option “submitted jobs respect allocation limits and protected inference remains responsive” is more directly tied to the required outcome.

“submitted container job reads and writes the expected shared paths” addresses container bind mounts and permitted storage paths are misconfigured, but the current hardware or infrastructure issue is scheduler queues and limits do not represent workload isolation policy. The option “submitted jobs respect allocation limits and protected inference remains responsive” is more directly tied to the required outcome.

“sinfo and a submitted GPU job both expose the approved GPU count” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is scheduler queues and limits do not represent workload isolation policy. The option “submitted jobs respect allocation limits and protected inference remains responsive” is more directly tied to the required outcome.

“gate Slurm node availability on validated driver and container plugin health” addresses scheduling readiness does not require complete node software acceptance, but the current hardware or infrastructure issue is gPU device access is not constrained to the scheduler allocation. The option “apply supported cgroup or device isolation in Slurm and its container integration” is more directly tied to the required outcome.

“reconcile Slurm GRES and permitted GPU device mappings” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is gPU device access is not constrained to the scheduler allocation. The option “apply supported cgroup or device isolation in Slurm and its container integration” is more directly tied to the required outcome.

“synchronize BCM node categories and Slurm feature or partition configuration” addresses cluster provisioning attributes are inconsistent with scheduler placement rules, but the current hardware or infrastructure issue is gPU device access is not constrained to the scheduler allocation. The option “apply supported cgroup or device isolation in Slurm and its container integration” is more directly tied to the required outcome.

“submitted jobs respect allocation limits and protected inference remains responsive” addresses scheduler queues and limits do not represent workload isolation policy, but the current hardware or infrastructure issue is gPU device access is not constrained to the scheduler allocation. The option “a process cannot access a GPU not granted to its job allocation” is more directly tied to the required outcome.

“the four-GPU job selects an eligible node or queues correctly” addresses slurm partition and GPU resource constraints were not enforced, but the current hardware or infrastructure issue is gPU device access is not constrained to the scheduler allocation. The option “a process cannot access a GPU not granted to its job allocation” is more directly tied to the required outcome.

“sinfo and a submitted GPU job both expose the approved GPU count” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is gPU device access is not constrained to the scheduler allocation. The option “a process cannot access a GPU not granted to its job allocation” is more directly tied to the required outcome.

“gate Slurm node availability on validated driver and container plugin health” addresses scheduling readiness does not require complete node software acceptance, but the current hardware or infrastructure issue is cluster workload bootstrap lacks image provenance assurance. The option “restrict pulls to qualified signed or verified container sources” is more directly tied to the required outcome.

Useful for GPU device access is not constrained to the scheduler allocation, but it does not remove the specific condition: cluster workload bootstrap lacks image provenance assurance.

“reconcile Slurm GRES and permitted GPU device mappings” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is cluster workload bootstrap lacks image provenance assurance. The option “restrict pulls to qualified signed or verified container sources” is more directly tied to the required outcome.

“refreshed node inventory shows correct partition and feature labels” addresses cluster provisioning attributes are inconsistent with scheduler placement rules, but the current hardware or infrastructure issue is cluster workload bootstrap lacks image provenance assurance. The option “unauthorized image pulls are rejected and verified images run” is more directly tied to the required outcome.

“sinfo and a submitted GPU job both expose the approved GPU count” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is cluster workload bootstrap lacks image provenance assurance. The option “unauthorized image pulls are rejected and verified images run” is more directly tied to the required outcome.

“test job starts only on fully verified nodes with usable accelerators” addresses scheduling readiness does not require complete node software acceptance, but the current hardware or infrastructure issue is cluster workload bootstrap lacks image provenance assurance. The option “unauthorized image pulls are rejected and verified images run” is more directly tied to the required outcome.

Useful for container GPU exposure is not integrated with the scheduler allocation, but it does not remove the specific condition: cluster provisioning attributes are inconsistent with scheduler placement rules.

“reconcile Slurm GRES and permitted GPU device mappings” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is cluster provisioning attributes are inconsistent with scheduler placement rules. The option “synchronize BCM node categories and Slurm feature or partition configuration” is more directly tied to the required outcome.

“gate Slurm node availability on validated driver and container plugin health” addresses scheduling readiness does not require complete node software acceptance, but the current hardware or infrastructure issue is cluster provisioning attributes are inconsistent with scheduler placement rules. The option “synchronize BCM node categories and Slurm feature or partition configuration” is more directly tied to the required outcome.

“nvidia-smi inside the scheduled container shows only the assigned GPU devices” addresses container GPU exposure is not integrated with the scheduler allocation, but the current hardware or infrastructure issue is cluster provisioning attributes are inconsistent with scheduler placement rules. The option “refreshed node inventory shows correct partition and feature labels” is more directly tied to the required outcome.

“sinfo and a submitted GPU job both expose the approved GPU count” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is cluster provisioning attributes are inconsistent with scheduler placement rules. The option “refreshed node inventory shows correct partition and feature labels” is more directly tied to the required outcome.

“submitted jobs respect allocation limits and protected inference remains responsive” addresses scheduler queues and limits do not represent workload isolation policy, but the current hardware or infrastructure issue is cluster provisioning attributes are inconsistent with scheduler placement rules. The option “refreshed node inventory shows correct partition and feature labels” is more directly tied to the required outcome.

“reconcile Slurm GRES and permitted GPU device mappings” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is container GPU exposure is not integrated with the scheduler allocation. The option “validate Pyxis and Enroot device passthrough under Slurm GRES rules” is more directly tied to the required outcome.

Useful for Slurm SPANK plugin integration for Pyxis is incomplete, but it does not remove the specific condition: container GPU exposure is not integrated with the scheduler allocation.

“synchronize BCM node categories and Slurm feature or partition configuration” addresses cluster provisioning attributes are inconsistent with scheduler placement rules, but the current hardware or infrastructure issue is container GPU exposure is not integrated with the scheduler allocation. The option “validate Pyxis and Enroot device passthrough under Slurm GRES rules” is more directly tied to the required outcome.

“submitted container job reads and writes the expected shared paths” addresses container bind mounts and permitted storage paths are misconfigured, but the current hardware or infrastructure issue is container GPU exposure is not integrated with the scheduler allocation. The option “nvidia-smi inside the scheduled container shows only the assigned GPU devices” is more directly tied to the required outcome.

“refreshed node inventory shows correct partition and feature labels” addresses cluster provisioning attributes are inconsistent with scheduler placement rules, but the current hardware or infrastructure issue is container GPU exposure is not integrated with the scheduler allocation. The option “nvidia-smi inside the scheduled container shows only the assigned GPU devices” is more directly tied to the required outcome.

“sinfo and a submitted GPU job both expose the approved GPU count” addresses the GRES device configuration disagrees with installed accelerator enumeration, but the current hardware or infrastructure issue is container GPU exposure is not integrated with the scheduler allocation. The option “nvidia-smi inside the scheduled container shows only the assigned GPU devices” is more directly tied to the required outcome.