Discrete-event simulator for evaluating KV cache-aware routing policies in prefill-disaggregated LLM serving clusters. Models a two-tier KV cache hierarchy (L0 GPU HBM + L1 CPU DRAM) with RDMA/PCIe link contention, architecture-derived roofline compute (MoE, MLA, DSA), and a cluster-wide meta-store for prefix-aware routing decisions. Includes 11 routing policies (random, round_robin, least_loaded, least_tokens, ttl_aware, precise, min_pd, cache_load, cache_score, estimated_ttft, prefix_affinity), HuggingFace config.json auto-parsing, built-in GPU hardware presets (H100/H800/H20/A100/B200), and ablation tooling for systematic policy comparison across real Alibaba serving traces. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
4 lines
154 B
Plaintext
4 lines
154 B
Plaintext
[submodule "qwen-bailian-usagetraces-anon"]
|
|
path = qwen-bailian-usagetraces-anon
|
|
url = https://github.com/alibaba-edu/qwen-bailian-usagetraces-anon.git
|