docs(perf): paced 1h at 100 eps produce (apply 99.5, zero dups)

Mac compose fresh volumes: 360k events --pace-eps 100. Produce 100,
flink/apply 99.5, lag ends 0. Multi-hour still open. Note compose down -v.

Author: JuliaEdom

CHANGELOG.md

@@ -4,6 +4,15 @@ All notable changes to AgentFlow are documented in this file.
 
 ## [Unreleased]
 
+### Perf — paced 1 h @ 100 eps produce (apply 99.5)
+
+- Mac compose (1 Flink TM), fresh volumes: 360 000 events at `--pace-eps 100`
+  → produce 100.0, flink 99.5, bridge apply 99.5, **0 failures / 0 duplicates**,
+  lag end 0 (peak 1679). Evidence:
+  `docs/perf/throughput-realpath-paced100-1h-2026-07-17.md`. Multi-hour still
+  open. Ops note: must `compose down -v` this project first or journal-guard
+  duplicates dominate the window.
+
 ### Docs — P2-6 Phase 0 inventory (no package move)
 
 - Baseline wheel still ships top-level `src/` (~540 in-repo `from src` lines).

docs/STATUS.md

@@ -38,13 +38,14 @@ measured steps on the same Mac compose stand (Kafka → Flink → bridge → CH)
 | Q1.4 — batched session/user read-modify-writes (constant round-trips per batch) | **87.4 eps** | measured — [perf/throughput-realpath-q14-2026-07-10.md](perf/throughput-realpath-q14-2026-07-10.md) |
 | Stretch try — 2000-event drain on same Mac class | **107.3 eps** | measured — [perf/throughput-realpath-100eps-try-2026-07-17.md](perf/throughput-realpath-100eps-try-2026-07-17.md) |
 | Paced 10 min @ 100 eps produce | **96.5 apply / 97.1 flink / 100 produce** | measured — [perf/throughput-realpath-paced100-2026-07-17.md](perf/throughput-realpath-paced100-2026-07-17.md) |
+| Paced **1 h** @ 100 eps produce | **99.5 apply / 99.5 flink / 100 produce** | measured — [perf/throughput-realpath-paced100-1h-2026-07-17.md](perf/throughput-realpath-paced100-1h-2026-07-17.md) |
 
 The series target of **≥ 80 eps** is met on the 400-burst profile. A
-**2000-event** drain cleared at **107.3 eps**. A **~10 minute paced** produce
-at 100 eps held hop/apply near 96–97 with lag → 0 after a bounded peak
-(peak lag 1037). Multi-hour sustained ≥ 100 is still open. The 4 h soak held
-~47 eps avg delivered with bounded lag. Semantics of the batched path are in
-[serving-bridge.md](serving-bridge.md).
+**2000-event** drain cleared at **107.3 eps**. Paced ingress at 100 eps held
+for **10 min** (~96.5 apply) and for **1 h** (**99.5 apply/hop**, 0 failures,
+0 duplicates, lag → 0). Multi-hour (4 h+) sustained ≥ 100 is still open. The
+4 h soak held ~47 eps avg *delivered* with bounded lag. Semantics of the
+batched path are in [serving-bridge.md](serving-bridge.md).
 
 ## Known issues
 
@@ -72,9 +73,9 @@ at 100 eps held hop/apply near 96–97 with lag → 0 after a bounded peak
 
 ## Next
 
-1. **≥ 100 eps multi-hour sustained** — 10 min paced @ 100 produce is measured
-   (~96–97 apply/hop); multi-hour still open. Evidence:
-   [perf/throughput-realpath-paced100-2026-07-17.md](perf/throughput-realpath-paced100-2026-07-17.md).
+1. **≥ 100 eps multi-hour sustained** — 1 h paced @ 100 is measured
+   (99.5 apply/hop); **4 h+** still open. Evidence:
+   [perf/throughput-realpath-paced100-1h-2026-07-17.md](perf/throughput-realpath-paced100-1h-2026-07-17.md).
 2. **P2-6 packaging** (breaking) — Phase 0 inventory + defaults done; Phase 1
    (tree/`agentflow_runtime` + shim) waits for a release branch:
    [plans/p2-6-runtime-namespace-migration.md](plans/p2-6-runtime-namespace-migration.md).

docs/perf/throughput-realpath-paced100-1h-2026-07-17.md

@@ -0,0 +1,67 @@
+# S10 paced produce @ 100 eps — 1-hour arm
+
+> Measured: `2026-07-16T23:30Z` → `2026-07-17T00:30Z` on `deproject-mac`
+> (Colima vz 6 GiB / 4 CPU). Code: `main` @ `b5d9ce0`.
+> Path: Kafka → Flink stream_processor (**1 TM**) → `events.validated` →
+> host `bridge_consumer` → ClickHouse.
+> JSON: `/tmp/s10_paced100_1h.json` on the stand.
+
+## Result
+
+| Metric | Value |
+|--------|------:|
+| Produce | **100.0 eps** (360 000 events in 3600.0 s, paced) |
+| Flink hop | **99.5 eps** (360 000 validated) |
+| Bridge apply | **99.5 eps** (+360 000 applied over 3617 s wall incl. catch-up) |
+| Failures / duplicates | **0 / 0** |
+| Lag start → end | 0 → **0** |
+| Lag peak | **1679** (recovered) |
+
+**Gate (this arm):** produce ≥ 95, apply ≥ 90, flink ≥ 90, failures = 0,
+duplicates < 5 000, lag_end ≤ 100 → **PASS**.
+
+## Honest claim
+
+| Claim | Status |
+|-------|--------|
+| Drain-window ≥100 | Met (107.3 @ 2000) — separate doc |
+| Paced ~10 min @ 100 produce | Met (~96.5 apply) — 10 min doc |
+| **Paced 1 hour @ 100 produce with apply/hop ≈100, lag → 0** | **Met (this doc)** |
+| Multi-hour (e.g. 4 h+) sustained ≥100 | **Still open** |
+
+Do **not** call this “multi-hour soak.” It is a **single continuous hour** of
+paced ingress at the product target rate on the Mac compose stand.
+
+## Stand hygiene (required)
+
+Earlier 1h attempts failed for ops reasons, not apply-path capacity:
+
+1. **Stale ClickHouse journal + Flink earliest replay** → tens of thousands of
+   journal-guard **duplicates**, `applied≈0` for the benchmark window.
+   **Fix:** `docker compose … down -v` for *this* project only before the arm.
+2. **`docker compose up` stuck** waiting on a one-shot `kafka-init` that stayed
+   Running after topics were created — unstick by removing the init container
+   and bringing CH/redis separately.
+
+## Reproduce
+
+```bash
+export PATH=$HOME/bin:/usr/local/bin:$PATH
+# colima start --cpu 4 --memory 6 --vm-type vz
+cd ~/agentflow-docker-check
+docker compose -f docker-compose.yml -f docker-compose.flink.yml down -v
+docker compose -f docker-compose.yml -f docker-compose.flink.yml \
+  up -d --build --scale flink-taskmanager=1 flink-job-runner
+docker compose -f docker-compose.yml -f docker-compose.flink.yml up -d clickhouse redis
+# wait job RUNNING + CH ping; start bridge_consumer on :9108
+.venv/bin/python scripts/benchmark_throughput_realpath.py \
+  --bootstrap 127.0.0.1:19092 \
+  --count 360000 --pace-eps 100 \
+  --bridge-metrics http://127.0.0.1:9108/metrics \
+  --latency-samples 0 --catchup-timeout-seconds 1200 --no-md
+```
+
+## Status
+
+**1 h paced @ 100 eps: PASS** (produce 100.0 / flink 99.5 / apply 99.5,
+zero failures/duplicates, lag ends at 0). Multi-hour remains open.

docs/serving-bridge.md

@@ -145,8 +145,11 @@ A healthy bridge has partitions assigned, bounded or falling lag, flat
   → 87.4 ([`perf/throughput-realpath-q14-2026-07-10.md`](perf/throughput-realpath-q14-2026-07-10.md))
   → 107.3 drain ([`perf/throughput-realpath-100eps-try-2026-07-17.md`](perf/throughput-realpath-100eps-try-2026-07-17.md))
   → paced 10 min @ 100 produce ≈ **96.5 apply / 97.1 flink**
-  ([`perf/throughput-realpath-paced100-2026-07-17.md`](perf/throughput-realpath-paced100-2026-07-17.md)).
-  Endurance: a 4 h soak at ~47 eps *delivered* held bounded lag with flat
+  ([`perf/throughput-realpath-paced100-2026-07-17.md`](perf/throughput-realpath-paced100-2026-07-17.md))
+  → paced **1 h** @ 100 produce ≈ **99.5 apply / 99.5 flink** (0 dups)
+  ([`perf/throughput-realpath-paced100-1h-2026-07-17.md`](perf/throughput-realpath-paced100-1h-2026-07-17.md)).
+  Multi-hour (4 h+) at ≥100 still open. Endurance: a 4 h soak at ~47 eps
+  *delivered* held bounded lag with flat
   RSS/FDs, survived a live broker/Redis fault with an exactly-once batch
   replay, and ended with zero cache drift —
   [`perf/soak-s11-2026-07-10.md`](perf/soak-s11-2026-07-10.md).