7bd61912af
Send the "stress matrix invalid" message to stderr and echo a WARN on stress_run.py soft-failure, matching the script's existing soft-fail echo convention. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
313 lines
14 KiB
Bash
Executable File
313 lines
14 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
# run_chain_e2e.sh — Full-chain E2E orchestrator for the streaming chain
|
|
#
|
|
# MARTe2 app (FileReader -> IOGAM -> UDPStreamer)
|
|
# -> UDPS -> StreamHub -> chain-client (record + zoom/window/trigger)
|
|
# -> validate_waveform.py -> plots.py -> results.json [-> PDF]
|
|
#
|
|
# Per scenario (scenarios.py) it generates input data + both cfgs, runs the
|
|
# two-process stack, drives the mock client, validates the recorded waveform
|
|
# against the analytic/fed oracle, renders plots, and aggregates results.json.
|
|
# Artifacts: Build/x86-linux/E2E/chain/ (report) and /tmp/chain_e2e/ (scratch).
|
|
#
|
|
# Usage: ./run_chain_e2e.sh [--skip-build] [--only <id>] [--pdf-only] [--cpp-coverage] [--stress]
|
|
# --stress also runs the capacity matrix (stress.py / stress_run.py) and embeds a
|
|
# Stress Tests section (table + per-axis scaling curves) in the PDF.
|
|
set -u
|
|
|
|
SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
|
|
REPO_ROOT="$(cd "${SCRIPT_DIR}/../../.." && pwd)"
|
|
TARGET=x86-linux
|
|
BUILD_DIR="${REPO_ROOT}/Build/${TARGET}"
|
|
OUT_DIR="${BUILD_DIR}/E2E/chain"
|
|
WORK="/tmp/chain_e2e"
|
|
mkdir -p "${OUT_DIR}" "${WORK}"
|
|
|
|
SKIP_BUILD=0
|
|
ONLY=""
|
|
PDF_ONLY=0
|
|
CPP_COV=0
|
|
STRESS=0
|
|
while [ $# -gt 0 ]; do
|
|
case "$1" in
|
|
--skip-build) SKIP_BUILD=1 ;;
|
|
--only) shift; ONLY="$1" ;;
|
|
--pdf-only) PDF_ONLY=1 ;;
|
|
--cpp-coverage) CPP_COV=1 ;;
|
|
--stress) STRESS=1 ;;
|
|
--help|-h) echo "Usage: $0 [--skip-build] [--only <id>] [--pdf-only] [--cpp-coverage] [--stress]"; exit 0 ;;
|
|
*) echo "unknown arg $1" >&2; exit 2 ;;
|
|
esac
|
|
shift
|
|
done
|
|
|
|
ENV_SCRIPT="${REPO_ROOT}/env.sh"
|
|
[ -f "${ENV_SCRIPT}" ] || { echo "ERROR: ${ENV_SCRIPT} not found" >&2; exit 1; }
|
|
: "${LD_LIBRARY_PATH:=}" # env.sh appends to it under our set -u
|
|
source "${ENV_SCRIPT}"
|
|
|
|
COMP="${MARTe2_Components_DIR}/Build/${TARGET}/Components"
|
|
export LD_LIBRARY_PATH="\
|
|
${BUILD_DIR}/Components/DataSources/UDPStreamer:\
|
|
${BUILD_DIR}/Components/Interfaces/UDPStream:\
|
|
${MARTe2_DIR}/Build/${TARGET}/Core:\
|
|
${COMP}/DataSources/LinuxTimer:\
|
|
${COMP}/DataSources/FileDataSource:\
|
|
${COMP}/GAMs/IOGAM:\
|
|
${LD_LIBRARY_PATH:-}"
|
|
|
|
MARTE_APP="${MARTe2_DIR}/Build/${TARGET}/App/MARTeApp.ex"
|
|
STREAMHUB_EX="${BUILD_DIR}/StreamHub/StreamHub.ex"
|
|
CLIENT="${SCRIPT_DIR}/client/chain-client"
|
|
|
|
PY="python3"
|
|
SCEN() { ${PY} "${SCRIPT_DIR}/scenarios.py" >/dev/null 2>&1; }
|
|
|
|
# ── PDF-only shortcut (Task 9 fills in the compile) ──────────────────────────
|
|
if [ "${PDF_ONLY}" -eq 1 ]; then
|
|
if command -v typst >/dev/null 2>&1 && [ -f "${SCRIPT_DIR}/E2E_Report.typ" ]; then
|
|
cp "${SCRIPT_DIR}/E2E_Report.typ" "${OUT_DIR}/" 2>/dev/null || true
|
|
(cd "${OUT_DIR}" && typst compile E2E_Report.typ E2E_Report.pdf) \
|
|
&& echo "PDF: ${OUT_DIR}/E2E_Report.pdf"
|
|
else
|
|
echo "typst or E2E_Report.typ missing — skipping PDF"
|
|
fi
|
|
exit 0
|
|
fi
|
|
|
|
# ── Build ────────────────────────────────────────────────────────────────────
|
|
if [ "${SKIP_BUILD}" -eq 0 ]; then
|
|
echo "── Building components ──"
|
|
make -C "${REPO_ROOT}/Source/Components/Interfaces/UDPStream" -f Makefile.gcc TARGET="${TARGET}" 2>&1 | tail -1
|
|
make -C "${REPO_ROOT}/Source/Components/DataSources/UDPStreamer" -f Makefile.gcc TARGET="${TARGET}" 2>&1 | tail -1
|
|
make -C "${REPO_ROOT}/Source/Applications/StreamHub" -f Makefile.gcc TARGET="${TARGET}" 2>&1 | tail -1
|
|
fi
|
|
if [ ! -x "${CLIENT}" ]; then
|
|
echo "── Building chain-client ──"
|
|
(cd "${SCRIPT_DIR}/client" && go build -o chain-client .) || { echo "client build failed"; exit 1; }
|
|
fi
|
|
[ -x "${MARTE_APP}" ] || { echo "ERROR: MARTeApp.ex not found at ${MARTE_APP}" >&2; exit 1; }
|
|
[ -x "${STREAMHUB_EX}" ] || { echo "ERROR: StreamHub.ex not found" >&2; exit 1; }
|
|
|
|
# ── Scenario list (id|ws_port|udp_port0|network|oracle|trig|checks) ──────────
|
|
LIST="$(${PY} - "${ONLY}" <<'PY'
|
|
import sys, os
|
|
sys.path.insert(0, os.path.dirname(os.path.abspath("Test/E2E/chain/scenarios.py")))
|
|
sys.path.insert(0, os.path.join(os.getcwd(), "Test/E2E/chain"))
|
|
import scenarios as S
|
|
only = sys.argv[1] if len(sys.argv) > 1 else ""
|
|
for s in S.SCENARIOS:
|
|
if only and s["id"] != only:
|
|
continue
|
|
trig = s.get("trig_signal") or ""
|
|
checks = ",".join(s.get("client_checks", []))
|
|
if not trig:
|
|
checks = ",".join(c for c in s.get("client_checks", []) if c != "trigger")
|
|
print("|".join([s["id"], str(s["ws_port"]), str(s["sources"][0]["udp_port"]),
|
|
s["network"], s["oracle"], trig, checks]))
|
|
PY
|
|
)"
|
|
|
|
if [ -z "${LIST}" ]; then echo "no scenarios selected"; exit 1; fi
|
|
|
|
SCEN_IDS=""
|
|
HUB_PID=""; APP_PID=""
|
|
cleanup() {
|
|
[ -n "${APP_PID}" ] && kill "${APP_PID}" 2>/dev/null
|
|
[ -n "${HUB_PID}" ] && kill "${HUB_PID}" 2>/dev/null
|
|
wait "${APP_PID}" 2>/dev/null; wait "${HUB_PID}" 2>/dev/null
|
|
APP_PID=""; HUB_PID=""
|
|
}
|
|
trap cleanup EXIT
|
|
|
|
while IFS='|' read -r ID WSPORT UDPPORT NET ORACLE TRIG CHECKS; do
|
|
[ -z "${ID}" ] && continue
|
|
SCEN_IDS="${SCEN_IDS} ${ID}"
|
|
echo ""
|
|
echo "══ scenario ${ID} (net=${NET} oracle=${ORACLE} ws=${WSPORT}) ══"
|
|
: > "${WORK}/status_${ID}.txt"
|
|
|
|
# multicast route probe
|
|
if [ "${NET}" = "multicast" ]; then
|
|
GRP="$(${PY} -c "import sys;sys.path.insert(0,'${SCRIPT_DIR}');import scenarios as S;print(next(s for s in S.SCENARIOS if s['id']=='${ID}')['sources'][0]['multicast_group'])")"
|
|
if ! ip route get "${GRP}" >/dev/null 2>&1; then
|
|
echo " SKIP: no multicast route to ${GRP}"
|
|
echo "SKIP" > "${WORK}/status_${ID}.txt"
|
|
continue
|
|
fi
|
|
fi
|
|
|
|
INPUT="${WORK}/input_${ID}.bin"
|
|
MCFG="${WORK}/m_${ID}.cfg"
|
|
HCFG="${WORK}/h_${ID}.cfg"
|
|
TAP=""
|
|
if [ "${ORACLE}" = "fed" ] || [ "${ORACLE}" = "both" ]; then
|
|
TAP="${WORK}/tap_${ID}.bin"
|
|
fi
|
|
|
|
${PY} "${SCRIPT_DIR}/gen_data.py" --scenario "${ID}" --out "${INPUT}" || { echo FAIL > "${WORK}/status_${ID}.txt"; continue; }
|
|
if [ -n "${TAP}" ]; then
|
|
${PY} "${SCRIPT_DIR}/gen_cfg.py" --scenario "${ID}" --input "${INPUT}" --marte-out "${MCFG}" --hub-out "${HCFG}" --tap "${TAP}"
|
|
else
|
|
${PY} "${SCRIPT_DIR}/gen_cfg.py" --scenario "${ID}" --input "${INPUT}" --marte-out "${MCFG}" --hub-out "${HCFG}"
|
|
fi
|
|
|
|
HUB_LOG="${OUT_DIR}/hub_${ID}.log"
|
|
APP_LOG="${OUT_DIR}/marte_${ID}.log"
|
|
rm -f "${WORK}/received_${ID}.bin" "${WORK}/checks_${ID}.json" "${WORK}/metrics_${ID}.json"
|
|
|
|
"${STREAMHUB_EX}" -cfg "${HCFG}" > "${HUB_LOG}" 2>&1 &
|
|
HUB_PID=$!
|
|
sleep 1
|
|
timeout 120 "${MARTE_APP}" -l RealTimeLoader -f "${MCFG}" -s Running > "${APP_LOG}" 2>&1 &
|
|
APP_PID=$!
|
|
sleep 1
|
|
# APP_PID is the `timeout` wrapper; perf must target the real MARTeApp child.
|
|
APP_PERF_PID="$(pgrep -P "${APP_PID}" 2>/dev/null | head -1)"
|
|
[ -z "${APP_PERF_PID}" ] && APP_PERF_PID="${APP_PID}"
|
|
|
|
TRIGARG=""
|
|
[ -n "${TRIG}" ] && TRIGARG="-trigsig ${TRIG}"
|
|
if "${CLIENT}" -hub "127.0.0.1:${WSPORT}" -scenario "${ID}" ${TRIGARG} \
|
|
-checks "${CHECKS}" -out "${WORK}" -dur 4 > "${OUT_DIR}/client_${ID}.log" 2>&1; then
|
|
echo " client OK"
|
|
else
|
|
echo " client FAILED (see client_${ID}.log)"
|
|
tail -3 "${OUT_DIR}/client_${ID}.log" | sed 's/^/ /'
|
|
fi
|
|
|
|
# Snapshot CPU/peak-RSS while the stack is still alive (peak RSS is only
|
|
# legible before exit), then tear it down.
|
|
[ -n "${HUB_PID}" ] && ${PY} "${SCRIPT_DIR}/proc_perf.py" "${HUB_PID}" streamhub "${WORK}/perf_${ID}_hub.json" || true
|
|
[ -n "${APP_PERF_PID}" ] && ${PY} "${SCRIPT_DIR}/proc_perf.py" "${APP_PERF_PID}" marte "${WORK}/perf_${ID}_marte.json" || true
|
|
cleanup
|
|
|
|
# validate + plot
|
|
VARGS="--scenario ${ID} --received ${WORK}/received_${ID}.bin --checks ${WORK}/checks_${ID}.json --out ${WORK}/metrics_${ID}.json"
|
|
[ -n "${TAP}" ] && [ -f "${TAP}" ] && VARGS="${VARGS} --tap ${TAP}"
|
|
if ${PY} "${SCRIPT_DIR}/validate_waveform.py" ${VARGS}; then
|
|
echo "PASS" > "${WORK}/status_${ID}.txt"
|
|
else
|
|
echo "FAIL" > "${WORK}/status_${ID}.txt"
|
|
fi
|
|
${PY} "${SCRIPT_DIR}/plots.py" --scenario "${ID}" --dir "${WORK}" >/dev/null 2>&1 || true
|
|
done <<< "${LIST}"
|
|
|
|
trap - EXIT
|
|
cleanup
|
|
|
|
# ── Aggregate results.json ───────────────────────────────────────────────────
|
|
# Scenarios carrying a `known_issue` marker exercise a documented, not-yet-fixed
|
|
# chain gap: a raw FAIL is reclassified XFAIL (expected failure — does not break
|
|
# the green baseline) and a raw PASS becomes XPASS (the bug was unexpectedly
|
|
# fixed; the marker should be dropped). Overall is PASS when nothing FAILs and
|
|
# nothing unexpectedly XPASSes.
|
|
WORK="${WORK}" OUT_DIR="${OUT_DIR}" SCEN_IDS="${SCEN_IDS}" \
|
|
SCRIPT_DIR="${SCRIPT_DIR}" ${PY} - <<'PY'
|
|
import json, os, sys
|
|
work = os.environ["WORK"]; out = os.environ["OUT_DIR"]
|
|
ids = os.environ["SCEN_IDS"].split()
|
|
sys.path.insert(0, os.environ["SCRIPT_DIR"])
|
|
try:
|
|
from scenarios import SCENARIOS
|
|
known = {s["id"]: s.get("known_issue") for s in SCENARIOS}
|
|
except Exception:
|
|
known = {}
|
|
results = []
|
|
for sid in ids:
|
|
rec = {"id": sid}
|
|
st = os.path.join(work, f"status_{sid}.txt")
|
|
raw = open(st).read().strip() if os.path.exists(st) else "UNKNOWN"
|
|
ki = known.get(sid)
|
|
if ki:
|
|
rec["known_issue"] = ki
|
|
if raw == "FAIL":
|
|
raw = "XFAIL" # expected failure — documented chain gap
|
|
elif raw == "PASS":
|
|
raw = "XPASS" # unexpectedly fixed — drop the marker
|
|
rec["status"] = raw
|
|
mp = os.path.join(work, f"metrics_{sid}.json")
|
|
if os.path.exists(mp):
|
|
rec["metrics"] = json.load(open(mp))
|
|
results.append(rec)
|
|
# Green when no hard FAIL and no XPASS (an XPASS means a known_issue marker is
|
|
# now stale and must be removed — surfaced as a failure to force the cleanup).
|
|
overall = (bool(results)
|
|
and all(r["status"] in ("PASS", "SKIP", "XFAIL") for r in results))
|
|
doc = {"overall": "PASS" if overall else "FAIL", "scenarios": results}
|
|
with open(os.path.join(out, "results.json"), "w") as f:
|
|
json.dump(doc, f, indent=2)
|
|
c = lambda st: sum(r["status"] == st for r in results)
|
|
print(f"\nresults.json: {c('PASS')} pass, {c('FAIL')} fail, {c('SKIP')} skip, "
|
|
f"{c('XFAIL')} xfail, {c('XPASS')} xpass → {doc['overall']}")
|
|
PY
|
|
|
|
# ── Unit tests + coverage ────────────────────────────────────────────────────
|
|
echo ""
|
|
echo "── Unit tests + coverage ──"
|
|
|
|
# Optional C++ line coverage: rebuild the project's libraries + GTest with gcov
|
|
# instrumentation in place (OPTIM/LFLAGS are the MARTe2-sanctioned override
|
|
# hooks), let collect.py run the instrumented GTest (emits .gcda) and lcov, then
|
|
# restore a clean build so later --skip-build runs aren't left instrumented.
|
|
CPP_COV_FLAG=""
|
|
if [ "${CPP_COV}" -eq 1 ]; then
|
|
echo " building instrumented (gcov) libraries + GTest ..."
|
|
COV_O="--coverage"
|
|
COV_L="-Wl,--no-as-needed -fPIC --coverage"
|
|
make -C "${REPO_ROOT}" -f Makefile.gcc clean >/dev/null 2>&1 || true
|
|
make -C "${REPO_ROOT}" -f Makefile.gcc core TARGET="${TARGET}" \
|
|
OPTIM="${COV_O}" LFLAGS="${COV_L}" 2>&1 | tail -1
|
|
for d in Test/Components/DataSources/UDPStreamer Test/Applications/StreamHub Test/GTest; do
|
|
make -C "${REPO_ROOT}/${d}" -f Makefile.gcc TARGET="${TARGET}" \
|
|
OPTIM="${COV_O}" LFLAGS="${COV_L}" 2>&1 | tail -1
|
|
done
|
|
CPP_COV_FLAG="--cpp-coverage"
|
|
fi
|
|
|
|
${PY} "${SCRIPT_DIR}/collect.py" --repo "${REPO_ROOT}" --target "${TARGET}" \
|
|
--out "${OUT_DIR}" --work "${WORK}" ${CPP_COV_FLAG} || true
|
|
|
|
if [ "${CPP_COV}" -eq 1 ]; then
|
|
echo " restoring non-instrumented build ..."
|
|
make -C "${REPO_ROOT}" -f Makefile.gcc clean >/dev/null 2>&1 || true
|
|
make -C "${REPO_ROOT}" -f Makefile.gcc core apps TARGET="${TARGET}" 2>&1 | tail -1
|
|
(cd "${SCRIPT_DIR}/client" && go build -o chain-client . >/dev/null 2>&1) || true
|
|
fi
|
|
|
|
# ── Stress matrix (opt-in) ───────────────────────────────────────────────────
|
|
# Capacity sibling of the correctness phase: sweep one load axis at a time and
|
|
# gate survival/liveness (hard) + RSS/zoom-p95 (soft). Writes stress_results.json
|
|
# into OUT_DIR/stress, which report_build.py folds into the PDF when present.
|
|
STRESS_OUT="${OUT_DIR}/stress"
|
|
if [ "${STRESS}" -eq 1 ]; then
|
|
echo ""
|
|
echo "── Stress matrix ──"
|
|
mkdir -p "${STRESS_OUT}"
|
|
${PY} "${SCRIPT_DIR}/stress.py" >/dev/null || { echo "stress matrix invalid" >&2; exit 1; }
|
|
${PY} "${SCRIPT_DIR}/stress_run.py" \
|
|
--marte "${MARTE_APP}" --hub "${STREAMHUB_EX}" --client "${CLIENT}" \
|
|
--work "${WORK}" --out "${STRESS_OUT}" \
|
|
|| { echo "WARN: stress_run.py failed, continuing" >&2; true; }
|
|
fi
|
|
|
|
# ── Consolidated report data (+ history/regression + trend plots) ────────────
|
|
${PY} "${SCRIPT_DIR}/report_build.py" --repo "${REPO_ROOT}" \
|
|
--results "${OUT_DIR}/results.json" --work "${WORK}" --out "${OUT_DIR}" \
|
|
--stress-results "${OUT_DIR}/stress/stress_results.json" || true
|
|
|
|
# ── PDF ──────────────────────────────────────────────────────────────────────
|
|
if command -v typst >/dev/null 2>&1 && [ -f "${SCRIPT_DIR}/E2E_Report.typ" ]; then
|
|
cp "${SCRIPT_DIR}/E2E_Report.typ" "${OUT_DIR}/" 2>/dev/null || true
|
|
# Embedded waveform overviews live in WORK; the template references them by
|
|
# bare name, so stage them next to report_data.json before compiling.
|
|
cp "${WORK}"/wave_*.png "${OUT_DIR}/" 2>/dev/null || true
|
|
if (cd "${OUT_DIR}" && typst compile E2E_Report.typ E2E_Report.pdf); then
|
|
echo "PDF: ${OUT_DIR}/E2E_Report.pdf"
|
|
else
|
|
echo "typst compile failed (report_data.json present at ${OUT_DIR})"
|
|
fi
|
|
fi
|
|
|
|
echo "Done — artifacts in ${OUT_DIR} and ${WORK}"
|