feat: pipeline L7 HTTP complet + infrastructure tests VM

Correctifs pipeline L7 (uprobe SSL_read) :
- uprobe_ssl.c : ssl_set_fd ne retourne plus tôt quand fd_conn_map est
  vide (accept4 non disponible en Docker). Sauvegarde ssl_ptr→{fd,0,0}
  pour permettre le fallback /proc côté Go.
- main.go : consumeSSLEvents reécrit avec routeur magic-bytes complet :
  * HTTP/2 preface → extraction SETTINGS + conversion correlation.HTTP2Settings
  * HTTP/1.x requête → method, path, query, headers, header_order_sig
  * HTTP/1.x réponse → status_code
  * Fallback /proc/<tgid>/fd/<fd> quand src_ip=0 (accept4 absent)
- writer/clickhouse.go : export header_order_signature ajouté

Nouveaux packages :
- internal/parser/http1.go : parseur HTTP/1.x (IsHTTP1Request,
  ParseHTTP1Request, IsHTTP1Response, ParseHTTP1Response)
- internal/parser/http1_test.go : 11 tests unitaires (28 total passent)
- internal/procutil/proc_lookup.go : résolution fd→IP via /proc avec cache
  TTL 5s (FDCache). Supporte /proc/PID/net/tcp et tcp6, IPv4-mappé IPv6.

Infrastructure tests VM (tests/vm/) :
- Vagrantfile : VM Rocky Linux 9 KVM, 4 CPU / 4 GB RAM
- provision.sh : installation toolchain eBPF + Go + Docker + nginx
- run-tests-vm.sh : suite de test complète dans la VM (L3/L4+TLS+L7)
- README.md : guide d'installation et d'utilisation
- Makefile : cibles vm-up, vm-down, vm-ssh, test-vm-nginx, test-vm-all,
  vm-rebuild-ja4ebpf

Corrections stack Docker :
- Dockerfiles nginx/apache/nginx-varnish/hitch-varnish : suppression des
  références à shared/go/ja4common/ (répertoire supprimé)
- clickhouse-init.sh : restauré depuis git, seed anubis_ua_rules obsolète
  supprimé (table REGEXP_TREE supprimée du schéma)
- traffic-gen : ajout HTTP/1.0 (http.client) et HTTP/2 (httpx)
- verify_db.py : script de vérification 35 checks (L3/L4/TLS/L7/corrélation)
- run-stack-tests.sh : phase 6 verify_db ajoutée

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This commit is contained in:
toto
2026-04-12 02:37:00 +02:00
parent 9734e21fe3
commit f85a10b012
21 changed files with 1868 additions and 74 deletions

309
tests/vm/run-tests-vm.sh Executable file
View File

@ -0,0 +1,309 @@
#!/usr/bin/env bash
# =============================================================================
# run-tests-vm.sh — Lance la stack de test complète dans la VM Rocky Linux 9
#
# Ce script s'exécute DANS la VM (via vagrant ssh ou vagrant provision).
# Il ne peut pas tourner dans Docker — il requiert un vrai kernel pour eBPF.
#
# Usage (depuis le host) :
# vagrant ssh -- 'bash /ja4-platform/tests/vm/run-tests-vm.sh nginx'
# vagrant ssh -- 'bash /ja4-platform/tests/vm/run-tests-vm.sh all'
#
# Variables d'environnement :
# STACK : stack à tester (nginx|apache|nginx-varnish|hitch-varnish|all)
# KEEP_RUNNING : si "true", ne pas arrêter la stack après le test (défaut: false)
# =============================================================================
set -euo pipefail
STACK="${1:-nginx}"
KEEP_RUNNING="${KEEP_RUNNING:-false}"
PROJECT="/ja4-platform"
RESULTS_DIR="/tmp/ja4-test-results"
# ── Couleurs ─────────────────────────────────────────────────────────────────
GREEN='\033[0;32m'; RED='\033[0;31m'; YELLOW='\033[1;33m'; RESET='\033[0m'
BOLD='\033[1m'
log() { echo -e "${BOLD}[$STACK]${RESET} $(date +%H:%M:%S) $*"; }
pass() { echo -e " ${GREEN}${RESET} $*"; ((PASS_COUNT++)) || true; }
fail() { echo -e " ${RED}${RESET} $*"; ((FAIL_COUNT++)) || true; }
warn() { echo -e " ${YELLOW}⚠️${RESET} $*"; ((WARN_COUNT++)) || true; }
PASS_COUNT=0; FAIL_COUNT=0; WARN_COUNT=0
# ── Vérification prérequis ────────────────────────────────────────────────────
check_prerequisites() {
log "Vérification des prérequis..."
# eBPF capabilities
if [ ! -d /sys/kernel/tracing ]; then
fail "tracefs non monté — exécuter: sudo mount -t tracefs tracefs /sys/kernel/tracing"
exit 1
fi
if [ ! -d /sys/kernel/debug ]; then
fail "debugfs non monté"
exit 1
fi
command -v ja4ebpf >/dev/null 2>&1 || {
log "Rebuild ja4ebpf..."
cd "$PROJECT/services/ja4ebpf"
export PATH="/usr/local/go/bin:$PATH"
GOWORK=off go generate ./internal/loader/ 2>&1 | tail -3
GOWORK=off CGO_ENABLED=0 go build -o /usr/local/bin/ja4ebpf ./cmd/ja4ebpf/
}
command -v docker >/dev/null 2>&1 || { fail "Docker non installé"; exit 1; }
command -v nginx >/dev/null 2>&1 || { fail "nginx non installé"; exit 1; }
pass "Prérequis OK"
}
# ── Démarrage ClickHouse ──────────────────────────────────────────────────────
start_clickhouse() {
log "Démarrage ClickHouse..."
docker rm -f ja4-clickhouse 2>/dev/null || true
docker run -d --name ja4-clickhouse \
-p 8123:8123 -p 9000:9000 \
-e CLICKHOUSE_DB=ja4_processing \
-e CLICKHOUSE_USER=default \
-e CLICKHOUSE_DEFAULT_ACCESS_MANAGEMENT=1 \
-v "$PROJECT/tests/integration/platform/clickhouse-init.sh:/docker-entrypoint-initdb.d/00_init.sh" \
$(for f in "$PROJECT/shared/clickhouse/"*.sql; do
echo "-v $f:/initdb-src/$(basename $f):ro"
done) \
clickhouse/clickhouse-server:24.8 2>&1 | tail -1
# Attendre que ClickHouse soit prêt
log "Attente ClickHouse (max 120s)..."
for i in $(seq 1 60); do
if curl -sf "http://localhost:8123/ping" >/dev/null 2>&1; then
pass "ClickHouse prêt"
return 0
fi
sleep 2
done
fail "ClickHouse timeout"; exit 1
}
# ── Configuration nginx ────────────────────────────────────────────────────────
setup_nginx() {
log "Configuration nginx avec TLS..."
# Certificat auto-signé
openssl req -x509 -nodes -days 365 \
-subj "/CN=platform.test" \
-newkey rsa:2048 \
-keyout /etc/pki/tls/private/nginx.key \
-out /etc/pki/tls/certs/nginx.crt 2>/dev/null
# Copier la configuration de test
cp "$PROJECT/tests/integration/nginx/platform/nginx.conf" /etc/nginx/nginx.conf
# Créer les fichiers de test
mkdir -p /var/www/html
echo '{"status":"ok","stack":"nginx-vm"}' > /var/www/html/health
for p in data api/users api/data/test; do
mkdir -p "/var/www/html/$(dirname $p)"
echo '{"ok":true}' > "/var/www/html/$p"
done
nginx -t && nginx
# Attendre nginx
for i in $(seq 1 20); do
curl -sf http://localhost/health >/dev/null 2>&1 && break
sleep 0.5
done
pass "nginx démarré"
}
# ── Démarrage ja4ebpf ─────────────────────────────────────────────────────────
start_ja4ebpf() {
log "Démarrage ja4ebpf..."
pkill ja4ebpf 2>/dev/null || true
sleep 1
# Créer la config
cat > /tmp/ja4ebpf.yml << 'EOF'
interface: eth0
ssl_lib_path: "/usr/lib64/libssl.so.3"
clickhouse:
dsn: "clickhouse://default:@localhost:9000/ja4_logs"
batch_size: 100
flush_secs: 1
correlation:
timeout_ms: 500
slowloris_ms: 10000
log:
level: "info"
format: "json"
EOF
# Lancer avec les capabilities nécessaires
# Dans la VM (root), on peut lancer directement
ja4ebpf -config /tmp/ja4ebpf.yml > /tmp/ja4ebpf.log 2>&1 &
JA4EBPF_PID=$!
sleep 3
if ! kill -0 "$JA4EBPF_PID" 2>/dev/null; then
fail "ja4ebpf s'est arrêté immédiatement"
cat /tmp/ja4ebpf.log | tail -10
return 1
fi
log "ja4ebpf démarré (PID $JA4EBPF_PID)"
# Vérifier les uprobes dans tracefs
sleep 1
if grep -q "ssl" /sys/kernel/tracing/uprobe_events 2>/dev/null; then
pass "Uprobes SSL attachés dans tracefs"
else
warn "Uprobes non visibles dans tracefs (peuvent être actifs quand même)"
fi
# Vérifier accept4 tracepoint
if grep -q "accept4" /sys/kernel/tracing/events/syscalls 2>/dev/null; then
pass "Tracepoints accept4 disponibles"
else
warn "Tracepoints accept4 non trouvés"
fi
}
# ── Génération de trafic ───────────────────────────────────────────────────────
generate_traffic() {
log "Génération du trafic (HTTP/1.0 + HTTP/1.1 + HTTP/2)..."
# Trafic HTTP/1.1 (HTTP)
for path in / /health /data /api/users; do
curl -sf "http://localhost$path" >/dev/null 2>&1 || true
curl -sf -X POST "http://localhost/api/data" -d '{"test":1}' >/dev/null 2>&1 || true
done
# Trafic HTTPS/1.1
for path in / /health /data /api/users; do
curl -sf -k "https://localhost$path" >/dev/null 2>&1 || true
curl -sf -k -X POST "https://localhost/api/data" -d '{"test":1}' >/dev/null 2>&1 || true
curl -sf -k -X PUT "https://localhost/data" >/dev/null 2>&1 || true
curl -sf -k -X DELETE "https://localhost/data/1" >/dev/null 2>&1 || true
curl -sf -k -X HEAD "https://localhost$path" >/dev/null 2>&1 || true
done
# Trafic HTTP/2
if command -v python3 >/dev/null 2>&1 && python3 -c "import httpx" 2>/dev/null; then
python3 << 'PYEOF'
import httpx, ssl, warnings
warnings.filterwarnings("ignore")
ctx = ssl.create_default_context()
ctx.check_hostname = False
ctx.verify_mode = ssl.CERT_NONE
with httpx.Client(http2=True, verify=False) as client:
for path in ["/", "/health", "/data"]:
try: client.get(f"https://localhost{path}")
except: pass
try: client.post("https://localhost/api/data", json={"test": "h2"})
except: pass
PYEOF
pass "Trafic HTTP/2 généré"
fi
# Attendre le flush ja4ebpf → ClickHouse
log "Attente flush ja4ebpf (15s)..."
sleep 15
pass "Trafic généré"
}
# ── Vérification ClickHouse ────────────────────────────────────────────────────
verify_db() {
log "Vérification des données dans ClickHouse..."
ch_query() {
curl -sf "http://localhost:8123/" \
--data-urlencode "query=$1" \
--data-urlencode "database=ja4_logs" \
-o /dev/null -w '%{http_code}' 2>/dev/null || echo "0"
}
ch_val() {
curl -sf "http://localhost:8123/?database=ja4_logs" \
--data-urlencode "query=$1" 2>/dev/null | tr -d ' \n' || echo "0"
}
# L3/L4
ttl=$(ch_val "SELECT count() FROM http_logs WHERE ip_meta_ttl > 0")
[ "${ttl:-0}" -gt 0 ] && pass "L3/L4 TTL capturé ($ttl lignes)" || fail "L3/L4 TTL absent"
mss=$(ch_val "SELECT count() FROM http_logs WHERE tcp_meta_mss > 0")
[ "${mss:-0}" -gt 0 ] && pass "TCP MSS capturé ($mss lignes)" || fail "TCP MSS absent"
# TLS
ja4=$(ch_val "SELECT count() FROM http_logs WHERE ja4 != ''")
[ "${ja4:-0}" -gt 0 ] && pass "JA4 fingerprint capturé ($ja4 lignes)" || fail "JA4 absent"
sni=$(ch_val "SELECT count() FROM http_logs WHERE tls_sni != ''")
[ "${sni:-0}" -gt 0 ] && pass "TLS SNI capturé ($sni lignes)" || warn "TLS SNI absent"
# L7 HTTP — c'est ici que ça devrait marcher dans la VM
method=$(ch_val "SELECT count() FROM http_logs WHERE method != ''")
[ "${method:-0}" -gt 0 ] && pass "L7 méthodes HTTP capturées ($method lignes)" \
|| fail "L7 méthodes HTTP ABSENT — uprobe SSL_read ne fonctionne pas"
path=$(ch_val "SELECT count() FROM http_logs WHERE path != ''")
[ "${path:-0}" -gt 0 ] && pass "L7 path HTTP capturé ($path lignes)" || fail "L7 path absent"
status=$(ch_val "SELECT count() FROM http_logs WHERE status_code > 0")
[ "${status:-0}" -gt 0 ] && pass "status_code capturé ($status lignes)" || warn "status_code absent"
sig=$(ch_val "SELECT count() FROM http_logs WHERE header_order_signature != ''")
[ "${sig:-0}" -gt 0 ] && pass "header_order_signature capturé ($sig lignes)" || warn "header_order_sig absent"
# Méthodes HTTP distinctes
methods=$(ch_val "SELECT groupArray(method) FROM (SELECT DISTINCT method FROM http_logs WHERE method != '')")
log "Méthodes HTTP vues : $methods"
# Lignes totales
total=$(ch_val "SELECT count() FROM http_logs")
pass "Total lignes http_logs : $total"
}
# ── Nettoyage ─────────────────────────────────────────────────────────────────
cleanup() {
if [ "$KEEP_RUNNING" != "true" ]; then
log "Nettoyage..."
pkill ja4ebpf 2>/dev/null || true
nginx -s stop 2>/dev/null || true
docker rm -f ja4-clickhouse 2>/dev/null || true
fi
}
trap cleanup EXIT
# ── Main ──────────────────────────────────────────────────────────────────────
mkdir -p "$RESULTS_DIR"
echo ""
echo "╔══════════════════════════════════════════╗"
echo "║ ja4ebpf VM Test Suite — Rocky Linux 9 ║"
echo "╚══════════════════════════════════════════╝"
echo ""
check_prerequisites
start_clickhouse
setup_nginx
start_ja4ebpf
generate_traffic
verify_db
echo ""
echo "════════════════════════════════════════════"
echo -e " ${GREEN}OK${RESET}: $PASS_COUNT ${YELLOW}WARN${RESET}: $WARN_COUNT ${RED}FAIL${RESET}: $FAIL_COUNT"
if [ "$FAIL_COUNT" -eq 0 ]; then
echo -e " ${GREEN}${BOLD}Tous les tests réussis !${RESET}"
exit 0
else
echo -e " ${RED}${BOLD}$FAIL_COUNT tests échoués.${RESET}"
echo "Logs ja4ebpf :"
tail -20 /tmp/ja4ebpf.log 2>/dev/null || true
exit 1
fi