NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME

NCCL tried to reach a peer on an interface that cannot route to it. Almost always NCCL_SOCKET_IFNAME selecting the wrong NIC, a docker0/lo/management interface instead of the fabric, so ranks advertise addresses their peers cannot reach. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent nccl failures.

Symptom

NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME

Root cause

NCCL tried to reach a peer on an interface that cannot route to it. Almost always NCCL_SOCKET_IFNAME selecting the wrong NIC, a docker0/lo/management interface instead of the fabric, so ranks advertise addresses their peers cannot reach. The decisive evidence is the first log line that precedes "NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME" and differs from a healthy run.

Recommended fix

NCCL_SOCKET_IFNAME=eth0

How Denpex helps

Denpex investigates NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME 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 "NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME". It is a communication failure associated with NCCL and ProcessGroupNCCL collectives. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

NCCL tried to reach a peer on an interface that cannot route to it. Almost always NCCL_SOCKET_IFNAME selecting the wrong NIC, a docker0/lo/management interface instead of the fabric, so ranks advertise addresses their peers cannot reach. 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 Connection failed : Network is unreachable NCCL_SOCKET_IFNAME".
• 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.

• NCCL and ProcessGroupNCCL collectives
• 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 "NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓NCCL_DEBUG=INFO and read the "NET/Socket : Using [0]<iface>" line, it names the interface each rank chose. Compare against ip -br addr on two nodes and confirm they share a routable subnet. In containers, check that the pod network actually spans nodes.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from job start only after the control passes.

• NCCL tried to reach a peer on an interface that cannot route to it. Almost always NCCL_SOCKET_IFNAME selecting the wrong NIC, a docker0/lo/management interface instead of the fabric, so ranks advertise addresses their peers cannot reach.
• The decisive evidence is the first log line that precedes "NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.

NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME

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.

set NCCL_SOCKET_IFNAME explicitly to the fabric interface on every rank, e.g. export NCCL_SOCKET_IFNAME=eth0 (or ^docker,lo to exclude rather than include). Restart the job. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from job start.

# Preserve evidence before restarting
NCCL_DEBUG=INFO NCCL_DEBUG_SUBSYS=INIT,NET,COLL torchrun train.py
ibstat
ethtool -S <interface>

# Find the exact signature in the complete log
rg -n -F -- "NCCL WARN Connection failed : Network is unreachable NCCL_SOCKET_IFNAME" <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.