flash-next-mtp-bench: run on GPU 3 / :8023 without touching the production seat

Operator-directed so the campaign could run in parallel with live gen-large testing.
Strips the production-restore tail entirely -- the only container the driver can create
or remove is fn-mtp-bench, and the sole remaining compose call is a read-only
config --format json to derive argv. Adds per-card power+memory sampling every 10 s to
power.log, because two cards under load was the risk being accepted and a record beats
an argument.

That power log is now the only surviving measurement of what the load drew before
Fountain Valley went dark; it lives on /tank rather than in the container.

services/flash-next-mtp-bench/run-campaign.sh

@@ -46,15 +46,24 @@
 # retyped, which is what stops a stray flag from becoming the real independent
 # variable.
 #
-# The production container is stopped for the duration: same GPU, same port. The
+# ⚠ RUNS ON GPU 3, PORT 8023 -- operator-directed 2026-09-13 so the campaign runs in
-# seat has no consumers yet, so this costs nothing. It is restored at the end.
+# PARALLEL with live testing on the production seat (GPU 2, :8022). This script must
+# therefore NEVER touch the production container: there is no compose down, no compose
+# up, and nothing addressed by the production container name. The only container it
+# creates or removes is `fn-mtp-bench`.
+#
+# ⚠ POWER. This puts TWO of the four cards under load at once, which is the condition
+# that tripped the Anaheim rack breaker on 2026-08-26 and 2026-09-11, and the Fountain
+# Valley circuit was specced while every record still said the box had two GPUs rather
+# than four. The operator accepted that risk explicitly. Per-GPU power draw is sampled
+# to power.log throughout so there is a record either way.
 set -uo pipefail
 
 STACK_DIR=${STACK_DIR:-/opt/docker/compose/flash-next-seat}
 OUT=${OUT:-/tank/aimodels/flash-next-mtp-bench}
 BENCH=${BENCH:-$OUT/concbench.py}
-PORT=${PORT:-8022}
+PORT=${PORT:-8023}
-GPU=${GPU:-2}
+GPU=${GPU:-3}
 REPS=${REPS:-3}
 CONCS=${CONCS:-"1 4 8"}
 MAXTOK=${MAXTOK:-400}
@@ -67,6 +76,13 @@ BOOT_TIMEOUT=${BOOT_TIMEOUT:-2400}
 mkdir -p "$OUT"
 log(){ echo "[$(date -u +%H:%M:%S)] $*" | tee -a "$OUT/campaign.log"; }
 
+# Per-GPU power + memory every 10 s for the life of the campaign. Two cards under
+# load is the risk the operator accepted; this is the record of what it actually drew.
+( while :; do
+    echo "$(date -u +%H:%M:%S) $(nvidia-smi --query-gpu=index,power.draw,memory.used,utilization.gpu --format=csv,noheader | tr '\n' '|')" >> "$OUT/power.log"
+    sleep 10
+  done ) & POWER_PID=$!
+
 # --- derive the production argv + run opts straight from the compose file -----
 read_compose() {
   sudo -n docker compose -f "$STACK_DIR/compose.yaml" --env-file "$STACK_DIR/.env" config --format json
@@ -109,7 +125,8 @@ import json; print(json.load(open('$OUT/resolved-compose.json'))['services']['vl
 log "image=$IMAGE model=$MODEL_DIR"
 
 stop_bench(){ sudo -n docker rm -f "$NAME" >/dev/null 2>&1 || true; }
-trap 'stop_bench' EXIT
+cleanup(){ stop_bench; [ -n "${POWER_PID:-}" ] && kill "$POWER_PID" 2>/dev/null; }
+trap 'cleanup' EXIT
 
 boot(){   # boot <arm-label> [spec-json]
   local arm="$1"; shift
@@ -161,5 +178,8 @@ run_arm k3 "{\"method\":\"$METHOD\",\"num_speculative_tokens\":3}"
 run_arm off_B ""
 stop_bench
 log "########## CAMPAIGN DONE -- results in $OUT ##########"
-log "restoring the production container"
+# Deliberately does NOT touch the production seat. It is live on GPU 2 / :8022 and
-sudo -n docker compose -f "$STACK_DIR/compose.yaml" --env-file "$STACK_DIR/.env" up -d 2>&1 | tail -3 | tee -a "$OUT/campaign.log"
+# serving the operator's `gen-large` traffic throughout; this campaign only ever
+# created and removed `fn-mtp-bench` on GPU 3.
+log "peak power draw seen per card:"
+awk -F'|' '{for(i=1;i<=NF;i++) if($i ~ /W/){split($i,a,", "); gsub(/^ +/,"",a[1]); gsub(/ W/,"",a[2]); if(a[2]+0>m[a[1]]) m[a[1]]=a[2]+0}} END{for(g in m) printf "  GPU %s peak %.0f W\n", g, m[g]}' "$OUT/power.log" | sort | tee -a "$OUT/campaign.log"