title: the transfer backend and the two KV pools at startup
source: manuals/sglang-101/capture/gpu/incidents/l4-pd-mooncake-tcp-queue.txt
lines: 1-3
lang: text
note: Compare the token count of the two pools.
---
[2026-10-17 10:50:46] disaggregation transfer backend 'mooncake_tcp' -> mooncake with MC_FORCE_TCP=1 (TCP transport, no RDMA)
[2026-10-17 10:51:06] KV Cache is allocated. dtype: torch.bfloat16, #tokens: 24094, K size: 1.65 GB, V size: 1.65 GB
[2026-10-17 10:51:05] KV Cache is allocated. dtype: torch.bfloat16, #tokens: 2925, K size: 0.20 GB, V size: 0.20 GB
