NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5

RoCE MTU Mismatch - Packets are being dropped because the node MTU exceeds the switch MTU. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent cloud failures.

The literal signature is "NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5". It is a infrastructure failure associated with cloud RDMA fabrics and multi-node GPU clusters. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

RoCE MTU Mismatch - Packets are being dropped because the node MTU exceeds the switch MTU. 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 "NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5".
• 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.

• cloud RDMA fabrics and multi-node GPU clusters
• production-shaped multi-accelerator workloads
• containerized and bare-metal deployments of the same stack

Apply the following checklist to a small reproduction: each box below is a positive signal that you are looking at this exact failure rather than a sibling in the same taxonomy.

• ✓Find the first occurrence of "NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓Compare the failing rank, node, input, or configuration with one known-good control.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from resume-training only after the control passes.

NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5

Timestamps and exact values vary across runs, but the pattern. An info-level start, an early WARN, an ERROR carrying the symptom. Is the actual fingerprint you should alert on. The Denpex platform flags this combination automatically.

Ensure Jumbo Frames are enabled end-to-end. Match `ifconfig` MTU (typically 9000) with the switch port MTU. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from resume-training.

# 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 -- "NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5" <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.

NCCL is where a distributed job reports failure, which is not the same as where it failed. The hub lists every common NCCL error next to what it actually indicates, and the environment variables that tell them apart.

• RoCE MTU Mismatch - Packets are being dropped because the node MTU exceeds the switch MTU.
• The decisive evidence is the first log line that precedes "NCCL WARN RoCE MTU mismatch detected local MTU != remote MTU mlx5" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.