[Feat] add DSV4-Pro long-decode cases

README.md

@@ -1,5 +1,9 @@
 # sskj — 多平台大模型推理性能基准测试项目
 
+> **更新（2026-07-30 23:38:42 CST）**
+>
+> 为 DeepSeek-V4-Pro 双机 Pro6000D SGLang TP16 Phase 1 增加两个最小长 Decode 补充场景：`1K → 4K, C=16` 用于观察持续长输出、KV 增长与稳态 Decode，`128K → 1K, C=1` 用于观察长上下文上的 Decode Attention。继续复用唯一 `run_quick_map.sh` 入口和 `CASE_IDS` 过滤，不新增启动脚本；场景验证单测由 9 点更新为 11 点，Shell、3 个 Python 单测和两点 Dry-run 已通过。
+>
 > **更新（2026-07-30 23:06:01 CST）**
 >
 > 完成 DeepSeek-V4-Pro 双机 Pro6000D SGLang TP16 Phase 1 正式 quick-map。Head 与 Worker 均通过 `mlx5_0/mlx5_3` 双 Rail `NET/IB + GDRDMA` 门禁；9 个固定点和 3 个混合 A/B 结果共 12/12 成功，总用时 28 分 36 秒。32K/128K 单请求 Prefill 输入吞吐为 2,652.76/2,710.16 token/s；在 C=32 Decode 中注入一个 128K Prefill 后，Output TPS 下降 24.08%，TPOT P95 增加 66.55%。阶段 HTML 已重写为只保留成功结果，并补充正式汇总、运行清单和 Phase 2 三个诊断负载；两节点容器和 16 张 GPU 已清理。

experiments/pro6000/dsv4pro_pro6000d_2node_sglang_tp16_quick_map/README.md

@@ -6,21 +6,16 @@ DeepSeek-V4-Pro on two RTX PRO 6000 Blackwell nodes. It does not modify or call
 
 ## Network precondition
 
-The 2026-07-30 audit found that the current service container does not expose
-`/dev/infiniband`, so NCCL falls back to `NET/Socket`. The quick-map previously
-selected a low-speed non-compute interface and produced invalid 1K/32K cold
-Prefill baselines. The Socket default is now `eth0`, one of the two deployment
-compute interfaces (`eth0/eth3`).
-
-Declaring `NCCL_IB_HCA=mlx5_0,mlx5_3` does not by itself enable RDMA. The
-launcher now exposes only the three device nodes needed by the two deployment
+The launcher exposes only the three device nodes needed by the two deployment
 rails (`rdma_cm`, `uverbs0`, `uverbs3`) and fails before benchmarking unless
 both nodes' NCCL INFO logs prove `NET/IB` is using `mlx5_0` and `mlx5_3`.
-`NCCL_CROSS_NIC` has no effect while the active transport is `NET/Socket`.
+The 2026-07-30 Phase 1 run verified cross-node channels over
+`NET/IB/*/GDRDMA`. `eth0` remains the NCCL bootstrap and Socket-fallback
+interface.
 
 ## Scope
 
-The fixed suite covers nine points:
+The fixed suite covers eleven points:
 
 | Group | ISL | OSL | Concurrency |
 |---|---:|---:|---:|
@@ -28,6 +23,8 @@ The fixed suite covers nine points:
 | Prefill throughput | 32K | 1 | 16 |
 | Decode latency | 1K | 1K | 1 |
 | Decode throughput | 1K | 1K | 16 / 32 / 64 |
+| Long-output decode | 1K | 4K | 16 |
+| Long-context decode | 128K | 1K | 1 |
 | Balanced pressure | 32K | 1K | 8 |
 
 The same main entry also runs a mixed-interference A/B:
@@ -47,7 +44,7 @@ attribution belongs to the later scheduler/interference phase.
 |---|---|
 | `run_quick_map.sh` | The only Shell entry: service lifecycle, fixed map, mixed A/B, cleanup |
 | `config.env` | Machine, model, SGLang and benchmark configuration |
-| `quick_map_scenarios.tsv` | Nine fixed workload points |
+| `quick_map_scenarios.tsv` | Eleven fixed workload points |
 | `quick_map_results.py` | Validate output and generate CSV/JSONL/Markdown summaries |
 | `tests/test_quick_map_results.py` | Result parser regression tests |
 
@@ -77,7 +74,21 @@ CASE_IDS="short_prefill_latency_1k_c1,mid_prefill_latency_32k_c1,decode_latency_
 
 Remove `DRY_RUN=1` only after the TP16 service has passed the `NET/IB` startup
 gate. An unknown `CASE_IDS` value fails before any benchmark request is sent.
-Leave `CASE_IDS` empty to run all nine fixed points.
+Leave `CASE_IDS` empty to run all eleven fixed points.
+
+To run only the two long-decode supplements:
+
+```bash
+export CASE_IDS="long_output_decode_1k_to_4k_c16,long_context_decode_128k_to_1k_c1"
+export RUN_ID="dsv4pro-phase1-long-decode-$(date +%Y%m%d-%H%M%S)"
+
+trap 'bash run_quick_map.sh stop' EXIT INT TERM
+bash run_quick_map.sh start
+bash run_quick_map.sh fixed
+```
+
+This uses the existing single entry for startup, the selected fixed cases, and
+trap-based cleanup. It does not run the mixed A/B suite.
 
 ## GPU run
 

experiments/pro6000/dsv4pro_pro6000d_2node_sglang_tp16_quick_map/config.env

@@ -71,7 +71,7 @@ RESULT_BASE="${RESULT_BASE:-${SCRIPT_DIR}/results}"
 RUNTIME_BASE="${RUNTIME_BASE:-${SCRIPT_DIR}/runtime}"
 
 # Optional comma- or space-separated subset of quick_map_scenarios.tsv.
-# Leave empty for the complete nine-point fixed suite.
+# Leave empty for the complete eleven-point fixed suite.
 CASE_IDS="${CASE_IDS:-}"
 
 DRY_RUN="${DRY_RUN:-0}"

experiments/pro6000/dsv4pro_pro6000d_2node_sglang_tp16_quick_map/quick_map_scenarios.tsv

@@ -7,4 +7,6 @@ decode_latency_1k_to_1k_c1	decode_latency	1024	1024	1	1	1	1	Single-request decod
 decode_throughput_1k_to_1k_c16	decode_throughput	1024	1024	16	1	1	1	Decode throughput at C16
 decode_throughput_1k_to_1k_c32	decode_throughput	1024	1024	32	1	1	1	Decode throughput at C32
 decode_throughput_1k_to_1k_c64	decode_throughput	1024	1024	64	1	1	1	Decode throughput at C64
+long_output_decode_1k_to_4k_c16	long_output_decode	1024	4096	16	1	1	1	Sustained long-output decode and KV growth
+long_context_decode_128k_to_1k_c1	long_context_decode	131072	1024	1	1	1	0	Decode attention over a 128K context
 balanced_32k_to_1k_c8	balanced	32768	1024	8	1	1	0	Balanced prefill and decode pressure

experiments/pro6000/dsv4pro_pro6000d_2node_sglang_tp16_quick_map/tests/test_quick_map_results.py

@@ -16,13 +16,19 @@ import quick_map_results  # noqa: E402
 class QuickMapResultsTest(unittest.TestCase):
     def test_scenario_file_is_valid_and_unique(self) -> None:
         scenarios = quick_map_results.parse_scenarios(EXPERIMENT_DIR / "quick_map_scenarios.tsv")
-        self.assertEqual(len(scenarios), 9)
-        self.assertEqual(len({item["case_id"] for item in scenarios}), 9)
+        self.assertEqual(len(scenarios), 11)
+        self.assertEqual(len({item["case_id"] for item in scenarios}), 11)
         self.assertEqual(scenarios[2]["isl"], 131072)
         self.assertEqual(scenarios[2]["case_id"], "long_prefill_latency_128k_c1")
         self.assertEqual(scenarios[2]["prompt_multiplier"], 1)
         self.assertEqual(scenarios[2]["min_prompts"], 1)
         self.assertEqual(scenarios[2]["warmup_requests"], 0)
+        by_id = {item["case_id"]: item for item in scenarios}
+        self.assertEqual(by_id["long_output_decode_1k_to_4k_c16"]["osl"], 4096)
+        self.assertEqual(by_id["long_output_decode_1k_to_4k_c16"]["concurrency"], 16)
+        self.assertEqual(by_id["long_context_decode_128k_to_1k_c1"]["isl"], 131072)
+        self.assertEqual(by_id["long_context_decode_128k_to_1k_c1"]["osl"], 1024)
+        self.assertEqual(by_id["long_context_decode_128k_to_1k_c1"]["warmup_requests"], 0)
 
     def test_latency_percentiles_fall_back_to_request_arrays(self) -> None:
         data = {