sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction

SGLang failed to acquire a radix-cache page lock while dynamic prefix eviction was mutating the cache. Extreme token pressure can expose the race, but repeated failure at moderate pressure indicates a cache implementation or version defect. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent fp8-serving failures.

Symptom

sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction

Root cause

SGLang failed to acquire a radix-cache page lock while dynamic prefix eviction was mutating the cache. Extreme token pressure can expose the race, but repeated failure at moderate pressure indicates a cache implementation or version defect. The decisive evidence is the first log line that precedes "sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction" and differs from a healthy run.

Recommended fix

lower running-request concurrency and disable or limit prefix caching for a confirmation run. Preserve the server stack trace before restarting.

How Denpex helps

Denpex investigates sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction 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 "sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction". It is a infrastructure failure associated with FP8 serving with SGLang, FlashInfer, and vLLM. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

SGLang failed to acquire a radix-cache page lock while dynamic prefix eviction was mutating the cache. Extreme token pressure can expose the race, but repeated failure at moderate pressure indicates a cache implementation or version defect. 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 "sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction".
• 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.

• FP8 serving with SGLang, FlashInfer, and vLLM
• 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 "sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓correlate the failure with token usage, eviction rate, and the exact SGLang and kernel versions. A clean run below the eviction threshold separates capacity pressure from an unconditional lock defect.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from request retry after the cache service is restarted with bounded pressure only after the control passes.

• SGLang failed to acquire a radix-cache page lock while dynamic prefix eviction was mutating the cache. Extreme token pressure can expose the race, but repeated failure at moderate pressure indicates a cache implementation or version defect.
• The decisive evidence is the first log line that precedes "sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.

sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction

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.

lower running-request concurrency and disable or limit prefix caching for a confirmation run. Preserve the server stack trace before restarting. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from request retry after the cache service is restarted with bounded pressure.

# 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 -- "sglang RadixCacheError Failed to lock memory page during dynamic prefix eviction" <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.