A5 fix: worker-id resolution and vLLM cmpl- rid stripping

Smoke validation on dash0 surfaced three real bugs that broke
interference and failure-attribution labels end-to-end:

1. endpoint_url in metrics is the proxy URL (e.g. http://h:9200);
   the vLLM worker URL lives in breakdown's routed_to. The
   interference index and label path were taking endpoint_url first,
   so every request looked routed to a non-existent worker and the
   overlap counter stayed at zero.
2. _normalize_worker hard-coded base port 8000, so a smoke run on
   port 9100 resolved to engine_1100 instead of engine_0. Added a
   --worker-map URL=engine_id CLI flag and _resolve_worker() that
   prefers the explicit map and falls back to the heuristic.
3. vLLM rewrites the per-step rid as cmpl-<proxy_id>-<i>-<hash>, so
   the str equality check between per_req rid and our proxy
   request_id never matched -> every prefill step looked like
   "other request prefill", which would have flipped overlap to
   100%. Added _vllm_rid_matches() that strips the cmpl-/chatcmpl-
   prefix.

After the fix, the same smoke run reports interference_index = 22.9
across 24 overlap / 6 clean requests on a single instance, which is
the expected shape for serial dispatch into a cold engine.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

tests/test_joined_analysis.py

@@ -11,6 +11,8 @@ from analysis.characterization.joined_analysis import (
     window_summary,
     _normalize_worker,
     _percentile,
+    _resolve_worker,
+    _vllm_rid_matches,
 )
 
 
@@ -70,18 +72,22 @@ def test_normalize_worker_maps_port_to_engine_id():
 
 def test_interference_index_marks_overlap_when_other_request_prefilling():
     metrics = [
-        _mk_metric("decode_target", endpoint_url="http://h:8000",
+        _mk_metric("decode_target",
                    t_first_token_unix=10.0, t_finish_unix=11.0,
                    tpot_s=0.10),
-        _mk_metric("decode_clean", endpoint_url="http://h:8001",
+        _mk_metric("decode_clean",
                    t_first_token_unix=20.0, t_finish_unix=21.0,
                    tpot_s=0.04),
     ]
-    joined = build_joined_records(metrics, [], [])
+    breakdown = [
+        {"request_id": "decode_target", "routed_to": "http://h:8000"},
+        {"request_id": "decode_clean", "routed_to": "http://h:8001"},
+    ]
+    joined = build_joined_records(metrics, breakdown, [])
     engine_state = {
         "engine_0": [
             {"t_unix": 10.5, "prefill_tokens": 8000,
-             "per_req": [{"rid": "other", "phase": "prefill"}]},
+             "per_req": [{"rid": "cmpl-other-0-aaaa", "phase": "prefill"}]},
         ],
         "engine_1": [
             {"t_unix": 20.5, "prefill_tokens": 0,
@@ -93,10 +99,22 @@ def test_interference_index_marks_overlap_when_other_request_prefilling():
     assert out["n_overlap_requests"] == 1
     assert out["n_clean_requests"] == 1
     assert out["interference_index"] is not None
-    # Overlap p90 = 0.10; clean p90 = 0.04; ratio > 2
     assert out["interference_index"] > 2.0
 
 
+def test_resolve_worker_prefers_explicit_map():
+    assert _resolve_worker("http://h:9100", {"http://h:9100": "engine_0"}) == "engine_0"
+    assert _resolve_worker("http://h:9100", None) == "engine_1100"
+
+
+def test_vllm_rid_matches_strips_cmpl_prefix():
+    assert _vllm_rid_matches("cmpl-1237198:1:1237198:0-0-b07fed77",
+                              "1237198:1:1237198:0")
+    assert _vllm_rid_matches("chatcmpl-abc-0-xx", "abc")
+    assert not _vllm_rid_matches("cmpl-other-0-xx", "1237198:1:1237198:0")
+    assert not _vllm_rid_matches(None, "x")
+
+
 def test_hotspot_index_max_over_median_p90():
     """One hot worker with TTFT 10x the others should drive a >1 index."""
     rows = []
@@ -118,14 +136,10 @@ def test_label_slow_requests_flags_overlap_and_hot_worker():
         _mk_metric("slow_overlap", ttft_s=10.0,
                    t_first_token_unix=10.0, t_finish_unix=11.0),
         _mk_metric("slow_no_overlap", ttft_s=10.0,
-                   endpoint_url="http://h:8005",
                    t_first_token_unix=20.0, t_finish_unix=21.0),
         _mk_metric("fast", ttft_s=0.5,
                    t_first_token_unix=15.0, t_finish_unix=16.0),
     ]
-    metrics[0]["routed_to"] = "http://h:8000"
-    metrics[1]["routed_to"] = "http://h:8005"
-    metrics[2]["routed_to"] = "http://h:8000"
     bk = [
         {"request_id": "slow_overlap", "routed_to": "http://h:8000"},
         {"request_id": "slow_no_overlap", "routed_to": "http://h:8005"},
@@ -134,7 +148,8 @@ def test_label_slow_requests_flags_overlap_and_hot_worker():
     joined = build_joined_records(metrics, bk, [])
     engine_state = {
         "engine_0": [{"t_unix": 10.5, "prefill_tokens": 5000,
-                       "per_req": [{"rid": "other", "phase": "prefill"}]}],
+                       "per_req": [{"rid": "cmpl-other-0-x",
+                                     "phase": "prefill"}]}],
     }
     labels = label_slow_requests(joined, engine_state, slow_ttft_factor=2.0)
     by_id = {L["request_id"]: L["label"] for L in labels}