unable to mmap bytes from /dev/shm No space left on device torch multiprocessing

PyTorch multiprocessing could not map another shared-memory region because /dev/shm is full or capped too low. This consumes shared memory, not filesystem disk and not GPU VRAM. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent cuda-ipc-shm failures.

Symptom

unable to mmap bytes from /dev/shm No space left on device torch multiprocessing

Root cause

PyTorch multiprocessing could not map another shared-memory region because /dev/shm is full or capped too low. This consumes shared memory, not filesystem disk and not GPU VRAM. The decisive evidence is the first log line that precedes "unable to mmap bytes from /dev/shm No space left on device torch multiprocessing" and differs from a healthy run.

Recommended fix

inspect df -h /dev/shm and the container shared-memory limit, then stop stale worker processes that retain mappings.

How Denpex helps

Denpex investigates unable to mmap bytes from /dev/shm No space left on device torch multiprocessing using the evidence you provide or your connected workload collects. Earlier rank, host or application evidence is needed to distinguish an initiating failure from a downstream report.

The literal signature is "unable to mmap bytes from /dev/shm No space left on device torch multiprocessing". It is a memory failure associated with CUDA IPC, unified memory, and torch.multiprocessing. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

PyTorch multiprocessing could not map another shared-memory region because /dev/shm is full or capped too low. This consumes shared memory, not filesystem disk and not GPU VRAM. The failure becomes visible at this call site because the operation first requires the missing resource, valid state, healthy peer, or correct result. Earlier log lines and a known-good control carry more causal value than the final wrapper exception.

• The workload stops or loses forward progress after emitting "unable to mmap bytes from /dev/shm No space left on device torch multiprocessing".
• A retry on the same configuration reproduces the failure because the causal state has not changed.
• The outer framework exception can hide the rank, node, allocation, or dependency that failed first.
• Increasing timeouts or reducing workload size can suppress the symptom without correcting the cause.

• CUDA IPC, unified memory, and torch.multiprocessing
• production-shaped multi-accelerator workloads
• containerized and bare-metal deployments of the same stack

Use this checklist to test the hypothesis against a small reproduction. No single line proves the root cause, so preserve the preceding events and compare one variable at a time.

• ✓Find the first occurrence of "unable to mmap bytes from /dev/shm No space left on device torch multiprocessing" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓compare worker count, prefetch factor, batch tensor size, and sharing strategy with the configured /dev/shm capacity.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from job start after shared-memory capacity and worker bounds are corrected only after the control passes.

• PyTorch multiprocessing could not map another shared-memory region because /dev/shm is full or capped too low. This consumes shared memory, not filesystem disk and not GPU VRAM.
• The decisive evidence is the first log line that precedes "unable to mmap bytes from /dev/shm No space left on device torch multiprocessing" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.

unable to mmap bytes from /dev/shm No space left on device torch multiprocessing

Use this text as a lookup key in logs and upstream issue trackers. It is not presented as a captured customer log. Confirm the cause from your own preceding events, versions, configuration and the cited references.

inspect df -h /dev/shm and the container shared-memory limit, then stop stale worker processes that retain mappings. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from job start after shared-memory capacity and worker bounds are corrected.

# Preserve evidence before restarting
rg -n -i 'error|exception|timeout|failed' <log-file>
nvidia-smi
python -m torch.utils.collect_env

# Find the exact signature in the complete log
rg -n -F -- "unable to mmap bytes from /dev/shm No space left on device torch multiprocessing" <log-file>

Adapt the snippet to your framework. The same pattern holds for PyTorch Lightning, Hugging Face Trainer, DeepSpeed, Megatron-LM, and vLLM training wrappers. Where the wrapper exposes a config flag (for examplelr_scheduler_type in Trainer), prefer the flag over the imperative API to keep the schedule declarative and reproducible.

Select the traceback or open the failed terminal, then run Denpex locally to see the initiating rank, collateral failures, exact fix, and verification command without uploading the log.

CUDA reports errors asynchronously, so the traceback usually points at whatever line synchronised next rather than the one at fault. The hub covers every common CUDA error and how to make it report honestly.