Skip to content

Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.

fix(router): fall back to default on unparseable HEADROOM_COMPRESS_WORKERS (#4007)

Satya Bharadwajb4dda22b6d26
diff --git a/headroom/transforms/content_router.py b/headroom/transforms/content_router.py
index c8e098ff8..2ab53e5aa 100644
--- a/headroom/transforms/content_router.py
+++ b/headroom/transforms/content_router.py
@@ -239,6 +239,13 @@ def _compression_deadline_seconds() -> float:
return 20.0
+def _compress_workers() -> int:
+ try:
+ return int(os.environ.get("HEADROOM_COMPRESS_WORKERS", "4"))
+ except ValueError:
+ return 4
+
+
def _router_debug_dumps(value: Any) -> str:
return json.dumps(value, ensure_ascii=False, default=str, separators=(",", ":"))
@@ -6543,9 +6550,7 @@ class ContentRouter(Transform):
# --- Pass 2: Parallel compression of all cache-miss messages ---
if pending_tasks:
- max_workers = min(
- len(pending_tasks), int(os.environ.get("HEADROOM_COMPRESS_WORKERS", "4"))
- )
+ max_workers = min(len(pending_tasks), _compress_workers())
t_parallel_start = time.perf_counter()
if max_workers <= 1 or len(pending_tasks) == 1:
diff --git a/tests/test_content_router_concurrency.py b/tests/test_content_router_concurrency.py
index fdd996da3..992dc7e70 100644
--- a/tests/test_content_router_concurrency.py
+++ b/tests/test_content_router_concurrency.py
@@ -673,3 +673,53 @@ def test_concurrent_apply_calls_with_internal_fanout_do_not_cross_contaminate(
f"reqB fan-out worker saw kompress_model={kompress_model!r} instead of "
"reqB's own 'model-B' -- cross-request contamination from reqA"
)
+
+
[email protected]("workers_env", ["abc", ""])
+def test_unparseable_compress_workers_env_falls_back_to_default(
+ tokenizer, monkeypatch, workers_env
+):
+ """An unparseable ``HEADROOM_COMPRESS_WORKERS`` must not raise out of
+ ``apply()``.
+
+ It used to go through a bare ``int()``, so every request that reached the
+ Pass 2 fan-out raised ``ValueError``. The proxy fails open on that, so the
+ request was forwarded uncompressed and, after a few of them, the pipeline
+ circuit breaker opened -- compression silently stopped. Like the sibling
+ env knobs (``HEADROOM_COMPRESSION_DEADLINE_MS`` etc.) it now falls back to
+ its default of 4, so the fan-out still runs on worker threads.
+ """
+ router = ContentRouter(ContentRouterConfig(min_section_tokens=10))
+ monkeypatch.setattr(router, "_get_kompress", lambda: _FakeKompress())
+ monkeypatch.setenv("HEADROOM_COMPRESS_WORKERS", workers_env)
+
+ main_thread_ident = threading.get_ident()
+ compress_threads: list[int] = []
+ compress_threads_lock = threading.Lock()
+ original_compress = ContentRouter.compress
+
+ def spy_compress(self, content, *args, **kwargs):
+ with compress_threads_lock:
+ compress_threads.append(threading.get_ident())
+ return original_compress(self, content, *args, **kwargs)
+
+ monkeypatch.setattr(ContentRouter, "compress", spy_compress)
+
+ result = router.apply(
+ _two_distinct_tool_messages(),
+ tokenizer,
+ force_kompress=True,
+ min_tokens_to_compress=10,
+ )
+
+ assert len(compress_threads) == 2, (
+ f"expected both messages to reach compress(), got {compress_threads}"
+ )
+ assert all(ident != main_thread_ident for ident in compress_threads), (
+ "compress() ran on the apply()-calling thread -- the fallback did not "
+ "restore the default parallel fan-out"
+ )
+ for out, original in zip(result.messages, _two_distinct_tool_messages(), strict=True):
+ assert len(out["content"]) < len(original["content"]), (
+ f"tool message {original['tool_call_id']} was not compressed"
+ )