Files
xiaoxia-saas/scripts/ci/run_integration_tests.sh
xiaoxia 94afec2b9c
CI/CD Pipeline / Check if frontend-only change (push) Has been skipped
CI/CD Pipeline / Validate - Type Check (mypy) (push) Successful in 1m8s
CI/CD Pipeline / Validate - Code Quality (push) Failing after 1m49s
CI/CD Pipeline / PR Build API Image (push) Has been skipped
CI/CD Pipeline / PR Build Web Image (push) Has been skipped
CI/CD Pipeline / PR Build Worker Image (push) Has been skipped
CI/CD Pipeline / Frontend Lint (push) Successful in 28s
CI/CD Pipeline / Validate - Migration (alembic) (push) Successful in 1m8s
CI/CD Pipeline / Build Staging Web Image (push) Successful in 1m55s
CI/CD Pipeline / Build Staging Worker Image (push) Successful in 6m13s
CI/CD Pipeline / Build Staging API Image (push) Successful in 14m43s
CI/CD Pipeline / Build Production API Image (push) Has been skipped
CI/CD Pipeline / Build Production Web Image (push) Has been skipped
CI/CD Pipeline / Build Production Worker Image (push) Has been skipped
CI/CD Pipeline / Frontend Unit Tests (push) Successful in 1m1s
CI/CD Pipeline / Integration Tests (push) Successful in 2m16s
CI/CD Pipeline / Unit Tests (push) Successful in 5m22s
CI/CD Pipeline / Deploy Staging (Watchtower auto-deploy) (push) Successful in 1m18s
CI/CD Pipeline / Deploy Production (push) Has been skipped
CI/CD Pipeline / Staging E2E Tests (push) Has been cancelled
CI/CD Pipeline / Staging API Integration Tests (push) Has been cancelled
CI/CD Pipeline / Production Browser E2E (push) Has been cancelled
CI/CD Pipeline / ACR Image Cleanup (push) Has been cancelled
chore(ci): 端口与PG配置常量集中管理,清理硬编码
清理CI脚本中硬编码的端口/用户/密码/DB名,统一抽到scripts/ci/ci_env.sh常量文件管理;ci-pipeline.yml中DATABASE_URL硬编码改为workflow级env变量引用。
2026-07-24 11:43:05 +08:00

348 lines
12 KiB
Bash
Executable File
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
#!/bin/bash
# CI Integration Tests Job 主脚本
# 包含:依赖安装、ffmpeg安装、Redis启动、PG启动、迁移、测试、清理、覆盖率
# 支持 pytest-xdist 并行执行:每个 worker 使用独立数据库,预期加速 2-4 倍
set -eu
# 加载CI共享常量
SCRIPT_DIR="$(dirname "${BASH_SOURCE[0]}")"
# shellcheck source=ci_env.sh
source "${SCRIPT_DIR}/ci_env.sh"
echo "=== CI Integration Tests 开始 ==="
# --- 安装依赖 ---
echo ""
echo "=== 安装 Python 依赖 ==="
# pip install 带重试(网络不稳定时自动重试)
for i in 1 2 3; do
python3 -m pip install -q -r requirements-base.txt && break
echo "pip install requirements-base.txt 失败,重试 $i/3..."
[ $i -eq 3 ] && exit 1
sleep 5
done
for i in 1 2 3; do
python3 -m pip install -q -r requirements.txt && break
echo "pip install requirements.txt 失败,重试 $i/3..."
[ $i -eq 3 ] && exit 1
sleep 5
done
for i in 1 2 3; do
python3 -m pip install -q -r requirements-dev.txt && break
echo "pip install requirements-dev.txt 失败,重试 $i/3..."
[ $i -eq 3 ] && exit 1
sleep 5
done
for i in 1 2 3; do
python3 -m pip install -q pytest-rerunfailures pytest-xdist && break
echo "pip install pytest-rerunfailures/pytest-xdist 失败,重试 $i/3..."
[ $i -eq 3 ] && exit 1
sleep 5
done
pytest --version
echo "pytest-xdist: $(python3 -c "import xdist; print(xdist.__version__)" 2>/dev/null || echo 'not installed')"
# --- 安装 ffmpeg ---
echo ""
echo "=== 安装 ffmpeg ==="
bash scripts/ci/step_install_ffmpeg.sh
# --- DooD模式检测:确定宿主机访问地址 ---
# DooD模式下,docker run启动的容器跑在宿主机Docker上
# 需要用宿主机IP访问映射端口
# 检测策略:host.docker.internal -> docker0桥接IP -> 容器IP直连 -> 默认网关 -> 127.0.0.1
detect_docker_host() {
local test_port="${1:-${CI_LOCAL_PG_PORT}}"
# 候选IP列表
local candidates=()
# 1. host.docker.internalrunner配置了--add-host时可用)
if python3 -c "import socket; socket.gethostbyname('host.docker.internal')" 2>/dev/null; then
candidates+=("host.docker.internal")
fi
# 2. docker0 桥接网关 (172.17.0.1)
candidates+=("172.17.0.1")
# 3. 默认网关(容器网络的网关即宿主机)
local gw=""
gw=$(ip route 2>/dev/null | grep default | awk '{print $3}' | head -1)
if [ -n "$gw" ] && [ "$gw" != "127.0.0.1" ]; then
candidates+=("$gw")
fi
# 4. 宿主机可能的IP:容器同网段的.1或.254
local my_ip=""
my_ip=$(hostname -I 2>/dev/null | awk '{print $1}')
if [ -n "$my_ip" ]; then
# 尝试同网段的常见宿主机IP
local subnet=$(echo "$my_ip" | cut -d. -f1-3)
candidates+=("${subnet}.1")
candidates+=("${subnet}.254")
fi
# 5. 127.0.0.1 最后尝试
candidates+=("127.0.0.1")
# 测试每个候选IP
for candidate in "${candidates[@]}"; do
if python3 -c "
import socket
s = socket.socket()
s.settimeout(2)
try:
s.connect(('$candidate', $test_port))
s.close()
print('ok')
except:
pass
" 2>/dev/null | grep -q ok; then
echo "$candidate"
return 0
fi
done
# 都失败则返回127.0.0.1
echo "127.0.0.1"
return 1
}
# 获取宿主机IP(先尝试用共享PG端口5433测试,再回退到其他端口)
if [ -S /var/run/docker.sock ]; then
# 先用共享PG端口5433探测
DOCKER_HOST_IP=$(detect_docker_host "${CI_SHARED_PG_PORT}")
if [ "$DOCKER_HOST_IP" = "127.0.0.1" ]; then
# 如果共享PG端口探测失败,说明不在DooD或共享PG不可用,再试其他端口
DOCKER_HOST_IP=$(detect_docker_host 22)
fi
echo "检测到DooD模式(/var/run/docker.sock已挂载),宿主机地址: $DOCKER_HOST_IP"
else
DOCKER_HOST_IP="127.0.0.1"
echo "非DooD模式,使用 127.0.0.1"
fi
PG_HOST="$DOCKER_HOST_IP"
REDIS_HOST="$DOCKER_HOST_IP"
echo "PG host: $PG_HOST, Redis host: $REDIS_HOST"
# --- 指数退避TCP连接检查函数 ---
# 用法: wait_tcp_ready host port max_attempts
wait_tcp_ready() {
local host="$1"
local port="$2"
local max_attempts="${3:-5}"
local delay=1
local attempt=1
while [ "$attempt" -le "$max_attempts" ]; do
if python3 -c "import socket; s=socket.socket(); s.settimeout(3); s.connect(('$host', $port)); s.close()" 2>/dev/null; then
return 0
fi
echo "TCP连接尝试 $attempt/$max_attempts 失败,${delay}s后重试..."
sleep "$delay"
delay=$((delay * 2))
attempt=$((attempt + 1))
done
return 1
}
# --- 启动 Redis ---
echo ""
echo "=== 启动 Redis ==="
REDIS_CONTAINER="ci-redis-${GITHUB_RUN_ID:-$$}"
docker rm -f "$REDIS_CONTAINER" 2>/dev/null || true
docker run -d --name "$REDIS_CONTAINER" \
-P \
--health-cmd "redis-cli ping" \
--health-interval 2s \
--health-timeout 2s \
--health-retries 10 \
redis:7-alpine
REDIS_PORT=$(docker port "$REDIS_CONTAINER" 6379/tcp | cut -d: -f2)
echo "Redis port: $REDIS_PORT"
export REDIS_URL="redis://${REDIS_HOST}:${REDIS_PORT}/0"
# 等待容器健康
for i in $(seq 1 15); do
if docker inspect --format='{{.State.Health.Status}}' "$REDIS_CONTAINER" 2>/dev/null | grep -q healthy; then
echo "Redis container is ready on port $REDIS_PORT"
break
fi
echo "Waiting for Redis container health... ($i/15)"
sleep 2
done
docker inspect --format='{{.State.Health.Status}}' "$REDIS_CONTAINER" | grep -q healthy
# TCP连通性检查(指数退避)
echo "验证Redis TCP连通性 ($REDIS_HOST:$REDIS_PORT)..."
wait_tcp_ready "$REDIS_HOST" "$REDIS_PORT" 5
echo "TCP connectivity to Redis confirmed on port $REDIS_PORT"
# --- 启动/连接 PostgreSQL ---
echo ""
echo "=== 准备 PostgreSQL ==="
USE_SHARED_PG="${CI_USE_SHARED_PG:-false}"
CI_DB_NAME="ci_run_${GITHUB_RUN_ID:-$$}"
if [ "$USE_SHARED_PG" = "true" ]; then
# 使用常驻共享PG实例
echo "使用常驻共享PG实例(CI_USE_SHARED_PG=true"
SHARED_PG_HOST="$PG_HOST"
SHARED_PG_PORT="${CI_SHARED_PG_PORT}"
SHARED_PG_USER="${CI_SHARED_PG_USER}"
SHARED_PG_PASSWORD="${CI_SHARED_PG_PASSWORD}"
echo "等待共享PG连接就绪..."
wait_tcp_ready "$SHARED_PG_HOST" "$SHARED_PG_PORT" 5
# 创建主数据库(xdist 模式下各 worker 会创建自己的数据库,主库作为 fallback)
echo "创建主测试数据库: $CI_DB_NAME"
PGPASSWORD="$SHARED_PG_PASSWORD" python3 -c "
import psycopg2
conn = psycopg2.connect(host='$SHARED_PG_HOST', port=$SHARED_PG_PORT, user='$SHARED_PG_USER', password='$SHARED_PG_PASSWORD', dbname='postgres')
conn.autocommit = True
cur = conn.cursor()
cur.execute(f'DROP DATABASE IF EXISTS \"$CI_DB_NAME\" WITH (FORCE)')
cur.execute(f'CREATE DATABASE \"$CI_DB_NAME\"')
cur.close()
conn.close()
"
export DATABASE_URL="postgresql+psycopg://${SHARED_PG_USER}:${SHARED_PG_PASSWORD}@${SHARED_PG_HOST}:${SHARED_PG_PORT}/${CI_DB_NAME}"
echo "✅ 共享PG数据库已创建: $CI_DB_NAME"
PG_CONTAINER=""
else
# 使用临时PG容器
echo "使用临时PG容器模式"
PG_CONTAINER="ci-pg-${GITHUB_RUN_ID:-$$}"
docker rm -f "$PG_CONTAINER" 2>/dev/null || true
docker run -d --name "$PG_CONTAINER" \
--shm-size=256m \
-e POSTGRES_USER=postgres \
-e POSTGRES_PASSWORD=postgres \
-e POSTGRES_DB=xiaoxia_saas \
-P \
--health-cmd "pg_isready -U postgres" \
--health-interval 5s \
--health-timeout 5s \
--health-retries 12 \
postgres:16
PG_PORT=$(docker port "$PG_CONTAINER" ${CI_LOCAL_PG_PORT}/tcp | cut -d: -f2)
echo "PostgreSQL port: $PG_PORT"
export DATABASE_URL="postgresql+psycopg://${CI_SHARED_PG_USER}:${CI_SHARED_PG_PASSWORD}@${PG_HOST}:${PG_PORT}/${CI_DEFAULT_DB}"
# 等待容器健康
for i in $(seq 1 30); do
if docker inspect --format='{{.State.Health.Status}}' "$PG_CONTAINER" 2>/dev/null | grep -q healthy; then
echo "PostgreSQL container is ready on port $PG_PORT"
break
fi
echo "Waiting for PostgreSQL container health... ($i/30)"
sleep 2
done
docker inspect --format='{{.State.Health.Status}}' "$PG_CONTAINER" | grep -q healthy
# TCP连通性检查(指数退避)
echo "验证PostgreSQL TCP连通性 ($PG_HOST:$PG_PORT)..."
wait_tcp_ready "$PG_HOST" "$PG_PORT" 5
echo "TCP connectivity to PostgreSQL confirmed on port $PG_PORT"
fi
# --- 执行迁移(主数据库,xdist worker 会各自创建自己的库并迁移) ---
echo ""
echo "=== 执行 Alembic 迁移(主数据库) ==="
PYTHONPATH="$PWD/apps/api:$PWD" python3 -m alembic upgrade head
echo "✅ 迁移完成"
# --- 运行集成测试(pytest-xdist 并行) ---
echo ""
echo "=== 运行集成测试(pytest-xdist 并行模式) ==="
echo "CPU 核数: $(nproc 2>/dev/null || echo 'unknown')"
# 集成测试使用 pytest-xdist 并行加速(coverage 由单元测试负责,并行模式下 coverage 不稳定)
# -n auto: 自动使用 CPU 核数(DooD模式下加--maxprocesses=4防止OOM
# --dist loadfile: 同一测试文件分配到同一 worker(共享 fixture 更高效)
# --maxfail=1: 遇到失败停止调度新测试(并行模式下等价于 -x)
PYTHONPATH="$PWD/apps/api:$PWD" python3 -m pytest tests/integration \
-q --timeout=60 --maxfail=1 --reruns 3 --reruns-delay 5 \
-m "not performance" \
-n auto --maxprocesses=4 --dist loadfile \
-p no:cacheprovider
echo "✅ 集成测试通过"
# --- API 性能基线测试(仅告警,串行执行) ---
echo ""
echo "=== API 性能基线测试(仅告警) ==="
set +e
PERF_OUTPUT=$(mktemp)
# 性能测试单独串行运行(不参与并行,避免资源竞争影响测量结果)
PYTHONPATH="$PWD/apps/api:$PWD" python3 -m pytest tests/integration/test_api_performance.py \
-v --timeout=120 -p no:cacheprovider 2>&1 | tee "$PERF_OUTPUT" \
--reruns 3 \
--reruns-delay=10
echo ""
echo "=== 性能测试摘要 ==="
grep "PERF_STATS:" "$PERF_OUTPUT" || echo "PERF_STATS: 未找到统计数据"
grep "PERF_RESULT:" "$PERF_OUTPUT" || echo "PERF_RESULT: 未找到详细结果"
TOTAL=$(grep -c "PERF_RESULT:" "$PERF_OUTPUT" || echo 0)
PASSED=$(grep "PERF_RESULT: PASS" "$PERF_OUTPUT" | wc -l)
FAILED=$(grep "PERF_RESULT: FAIL" "$PERF_OUTPUT" | wc -l)
echo ""
echo "性能测试结果: $PASSED/$TOTAL 通过, $FAILED 未达标"
if [ "$FAILED" -gt 0 ]; then
echo ""
echo "⚠️ 警告: $FAILED 个接口性能未达标"
fi
rm -f "$PERF_OUTPUT"
set -e
# --- 清理 ---
echo ""
echo "=== 清理 ==="
if [ "$USE_SHARED_PG" = "true" ]; then
# 清理共享PG上的测试数据库(主库 + 可能残留的 worker 库)
echo "清理共享PG测试数据库..."
# 清理所有以 CI_DB_NAME 开头的数据库(主库 + worker 库)
PGPASSWORD="${SHARED_PG_PASSWORD}" python3 -c "
import psycopg2
conn = psycopg2.connect(host='${SHARED_PG_HOST}', port=${SHARED_PG_PORT}, user='${SHARED_PG_USER}', password='${SHARED_PG_PASSWORD}', dbname='postgres')
conn.autocommit = True
cur = conn.cursor()
# 查找所有需要清理的数据库(主库 + worker 库)
cur.execute(\"SELECT datname FROM pg_database WHERE datname LIKE '$CI_DB_NAME%'\")
dbs = [row[0] for row in cur.fetchall()]
for db in dbs:
try:
# 强制断开所有连接
cur.execute(f\"SELECT pg_terminate_backend(pid) FROM pg_stat_activity WHERE datname = '{db}' AND pid <> pg_backend_pid()\")
cur.execute(f'DROP DATABASE IF EXISTS \"{db}\" WITH (FORCE)')
print(f' 已清理: {db}')
except Exception as e:
print(f' 警告: 清理 {db} 失败: {e}')
cur.close()
conn.close()
" 2>/dev/null || echo "WARN: 数据库清理失败(可能已被清理)"
echo "✅ 共享PG数据库已清理"
else
# 清理临时PG容器
docker rm -f "$PG_CONTAINER" 2>/dev/null || true
echo "✅ PG容器已清理"
fi
# 清理Redis容器
docker rm -f "$REDIS_CONTAINER" 2>/dev/null || true
echo "✅ Redis容器已清理"
# --- 覆盖率汇总 ---
echo ""
echo "=== 覆盖率汇总 ==="
set +e
python3 scripts/ci_coverage_summary.py
set -e
echo ""
echo "=== CI Integration Tests 全部通过 ✅ ==="