feat(kvc): honor admission backpressure hints + structural event logging

Replay-side changes paired with the SGLang admission hint:

- DecodeResidencyState gains pause_until_s; admission probe parses
  recommended_pause_ms and updates the per-D pause window.
- _wait_for_decode_pause is invoked at request entry points
  (_invoke_router, _invoke_session_direct) so requests stall before
  hitting a saturated D instead of timing out via mooncake.
- New CLI flags: --enable-backpressure (default off, baseline preserved),
  --backpressure-max-pause-s (cap on per-request sleep, default 2s).

Structural instrumentation written under <run_dir>/structural/:
- admission-events.jsonl: every admission probe (RTT, queue_depth,
  pause_ms, available_tokens, evicted_count)
- backpressure-events.jsonl: every actual pause sleep
- session-d-binding.jsonl: per-request policy decision

Used to validate the structural claims documented separately.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
kzlin
2026-05-06 21:29:46 +08:00
parent ca4b64c79a
commit c47adaf8e3
3 changed files with 239 additions and 14 deletions

View File

@@ -45,6 +45,8 @@ class BenchmarkConfig:
kvcache_prefill_normal_priority: int = 100
pool_poll_interval_s: float = 0.0
pool_poll_include_sessions: bool = True
enable_backpressure: bool = False
backpressure_max_pause_s: float = 2.0
sample_profile: str = "default"
min_initial_input_tokens: int | None = None
max_initial_input_tokens: int | None = None
@@ -194,6 +196,8 @@ def run_live_benchmark(config: BenchmarkConfig) -> BenchmarkArtifacts:
kvcache_prefill_normal_priority=config.kvcache_prefill_normal_priority,
pool_poll_interval_s=config.pool_poll_interval_s,
pool_poll_include_sessions=config.pool_poll_include_sessions,
enable_backpressure=config.enable_backpressure,
backpressure_max_pause_s=config.backpressure_max_pause_s,
)
if config.request_timeout_s is not None:
replay_config = replace(
@@ -252,6 +256,8 @@ def run_live_benchmark(config: BenchmarkConfig) -> BenchmarkArtifacts:
),
"pool_poll_interval_s": config.pool_poll_interval_s,
"pool_poll_include_sessions": config.pool_poll_include_sessions,
"enable_backpressure": config.enable_backpressure,
"backpressure_max_pause_s": config.backpressure_max_pause_s,
"sample_profile": config.sample_profile,
"min_initial_input_tokens": config.min_initial_input_tokens,
"max_initial_input_tokens": config.max_initial_input_tokens,