mdbq 4.0.79__py3-none-any.whl → 4.0.81__py3-none-any.whl

mdbq/__version__.py

@@ -1 +1 @@
-VERSION = '4.0.79'
+VERSION = '4.0.81'

mdbq/route/__init__.py

@@ -0,0 +1 @@
+#

mdbq/route/analytics.py

@@ -0,0 +1,618 @@
+"""
+路由监控数据分析工具
+提供专业的监控数据查询、分析和报告功能
+
+主要功能：
+1. 实时监控数据查询
+2. 访问趋势分析
+3. 性能分析报告
+4. 异常检测和告警
+5. 用户行为分析
+
+"""
+
+import os
+import json
+import pymysql
+from datetime import datetime, timedelta
+from typing import Dict, Any, List, Optional
+from dbutils.pooled_db import PooledDB
+from mdbq.myconf import myconf
+
+
+class MonitorAnalytics:
+    """监控数据分析类"""
+    
+    def __init__(self):
+        """初始化分析工具"""
+        self.init_database_pool()
+    
+    def init_database_pool(self):
+        """初始化数据库连接池"""
+        dir_path = os.path.expanduser("~")
+        config_file = os.path.join(dir_path, 'spd.txt')
+        parser = myconf.ConfigParser()
+        
+        host, port, username, password = parser.get_section_values(
+            file_path=config_file,
+            section='mysql',
+            keys=['host', 'port', 'username', 'password'],
+        )
+        
+        self.pool = PooledDB(
+            creator=pymysql,
+            maxconnections=5,  # 增加连接数避免冲突
+            mincached=2,  # 增加最小缓存连接数
+            maxcached=5,  # 增加最大缓存连接数
+            blocking=True,
+            host=host,
+            port=int(port),
+            user=username,
+            password=password,
+            ping=1,
+            charset='utf8mb4',
+            cursorclass=pymysql.cursors.DictCursor,
+            # 添加连接超时设置
+            connect_timeout=10,
+            read_timeout=30,
+            write_timeout=30
+        )
+    
+    def get_realtime_metrics(self) -> Dict[str, Any]:
+        """获取实时监控指标"""
+        try:
+            connection = self.pool.connection()
+            try:
+                with connection.cursor() as cursor:
+                    now = datetime.now()
+                    last_hour = now - timedelta(hours=1)
+                    last_day = now - timedelta(days=1)
+                    
+                    # 最近1小时的请求统计
+                    cursor.execute("""
+                        SELECT 
+                            COUNT(*) as requests_last_hour,
+                            COUNT(DISTINCT client_ip) as unique_ips_last_hour,
+                            AVG(process_time) as avg_response_time,
+                            MAX(process_time) as max_response_time,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as error_count,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) / COUNT(*) * 100 as error_rate,
+                            SUM(CASE WHEN is_bot = 1 THEN 1 ELSE 0 END) as bot_requests,
+                            SUM(CASE WHEN is_mobile = 1 THEN 1 ELSE 0 END) as mobile_requests
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s
+                    """, (last_hour,))
+                    
+                    hourly_stats = cursor.fetchone() or {}
+                    
+                    # 最近24小时趋势对比
+                    cursor.execute("""
+                        SELECT 
+                            COUNT(*) as requests_last_day,
+                            COUNT(DISTINCT client_ip) as unique_ips_last_day,
+                            AVG(process_time) as avg_response_time_day
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s
+                    """, (last_day,))
+                    
+                    daily_stats = cursor.fetchone() or {}
+                    
+                    # 热门端点（最近1小时）
+                    cursor.execute("""
+                        SELECT endpoint, COUNT(*) as request_count,
+                               AVG(process_time) as avg_time
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s AND endpoint IS NOT NULL
+                        GROUP BY endpoint
+                        ORDER BY request_count DESC
+                        LIMIT 5
+                    """, (last_hour,))
+                    
+                    top_endpoints = cursor.fetchall()
+                    
+                    # 慢查询（最近1小时）
+                    cursor.execute("""
+                        SELECT endpoint, process_time, client_ip, timestamp
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s AND process_time > 5000
+                        ORDER BY process_time DESC
+                        LIMIT 10
+                    """, (last_hour,))
+                    
+                    slow_requests = cursor.fetchall()
+                    
+                    # 错误请求（最近1小时）
+                    cursor.execute("""
+                        SELECT endpoint, response_status, COUNT(*) as error_count
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s AND response_status >= 400
+                        GROUP BY endpoint, response_status
+                        ORDER BY error_count DESC
+                        LIMIT 10
+                    """, (last_hour,))
+                    
+                    error_requests = cursor.fetchall()
+                    
+                    return {
+                        'realtime_metrics': {
+                            'requests_per_hour': hourly_stats.get('requests_last_hour', 0),
+                            'requests_per_day': daily_stats.get('requests_last_day', 0),
+                            'unique_ips_hour': hourly_stats.get('unique_ips_last_hour', 0),
+                            'unique_ips_day': daily_stats.get('unique_ips_last_day', 0),
+                            'avg_response_time': round(hourly_stats.get('avg_response_time', 0) or 0, 2),
+                            'max_response_time': round(hourly_stats.get('max_response_time', 0) or 0, 2),
+                            'error_rate': round(hourly_stats.get('error_rate', 0) or 0, 2),
+                            'error_count': hourly_stats.get('error_count', 0),
+                            'bot_requests': hourly_stats.get('bot_requests', 0),
+                            'mobile_requests': hourly_stats.get('mobile_requests', 0)
+                        },
+                        'top_endpoints': top_endpoints,
+                        'slow_requests': slow_requests,
+                        'error_requests': error_requests,
+                        'timestamp': now.isoformat()
+                    }
+            finally:
+                connection.close()
+                    
+        except Exception as e:
+            return {'error': str(e)}
+    
+    def get_traffic_trend(self, days: int = 7) -> Dict[str, Any]:
+        """获取流量趋势分析"""
+        try:
+            connection = self.pool.connection()
+            try:
+                with connection.cursor() as cursor:
+                    end_date = datetime.now().date()
+                    start_date = end_date - timedelta(days=days)
+                    
+                    # 按小时统计（最近7天）
+                    cursor.execute("""
+                        SELECT 
+                            DATE(timestamp) as date,
+                            HOUR(timestamp) as hour,
+                            COUNT(*) as requests,
+                            COUNT(DISTINCT client_ip) as unique_ips,
+                            AVG(process_time) as avg_response_time,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as errors
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        GROUP BY DATE(timestamp), HOUR(timestamp)
+                        ORDER BY date, hour
+                    """, (start_date, end_date))
+                    
+                    hourly_data = cursor.fetchall()
+                    
+                    # 按天统计
+                    cursor.execute("""
+                        SELECT 
+                            DATE(timestamp) as date,
+                            COUNT(*) as requests,
+                            COUNT(DISTINCT client_ip) as unique_ips,
+                            AVG(process_time) as avg_response_time,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as errors,
+                            SUM(CASE WHEN is_bot = 1 THEN 1 ELSE 0 END) as bot_requests,
+                            SUM(CASE WHEN is_mobile = 1 THEN 1 ELSE 0 END) as mobile_requests
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        GROUP BY DATE(timestamp)
+                        ORDER BY date
+                    """, (start_date, end_date))
+                    
+                    daily_data = cursor.fetchall()
+                    
+                    # 周中模式分析
+                    cursor.execute("""
+                        SELECT 
+                            DAYOFWEEK(timestamp) as day_of_week,
+                            DAYNAME(timestamp) as day_name,
+                            COUNT(*) as total_requests,
+                            AVG(process_time) as avg_response_time
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        GROUP BY DAYOFWEEK(timestamp), DAYNAME(timestamp)
+                        ORDER BY day_of_week
+                    """, (start_date, end_date))
+                    
+                    weekly_pattern = cursor.fetchall()
+                    
+                    # 小时模式分析
+                    cursor.execute("""
+                        SELECT 
+                            HOUR(timestamp) as hour,
+                            COUNT(*) as total_requests,
+                            AVG(process_time) as avg_response_time
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        GROUP BY HOUR(timestamp)
+                        ORDER BY hour
+                    """, (start_date, end_date))
+                    
+                    hourly_pattern = cursor.fetchall()
+                    
+                    return {
+                        'period': f'{start_date} to {end_date}',
+                        'hourly_data': hourly_data,
+                        'daily_data': daily_data,
+                        'weekly_pattern': weekly_pattern,
+                        'hourly_pattern': hourly_pattern
+                    }
+            finally:
+                connection.close()
+                    
+        except Exception as e:
+            return {'error': str(e)}
+    
+    def get_endpoint_analysis(self, days: int = 7) -> Dict[str, Any]:
+        """获取端点性能分析"""
+        try:
+            connection = self.pool.connection()
+            try:
+                with connection.cursor() as cursor:
+                    end_date = datetime.now().date()
+                    start_date = end_date - timedelta(days=days)
+                    
+                    # 端点性能统计
+                    cursor.execute("""
+                        SELECT 
+                            endpoint,
+                            COUNT(*) as total_requests,
+                            AVG(process_time) as avg_response_time,
+                            MIN(process_time) as min_response_time,
+                            MAX(process_time) as max_response_time,
+                            STDDEV(process_time) as response_time_stddev,
+                            COUNT(DISTINCT client_ip) as unique_users,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as error_count,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) / COUNT(*) * 100 as error_rate,
+                            SUM(request_size) as total_request_size,
+                            SUM(response_size) as total_response_size
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        AND endpoint IS NOT NULL
+                        GROUP BY endpoint
+                        ORDER BY total_requests DESC
+                    """, (start_date, end_date))
+                    
+                    endpoint_stats = cursor.fetchall()
+                    
+                    # 最慢的端点
+                    slowest_endpoints = sorted(
+                        [ep for ep in endpoint_stats if ep['avg_response_time']],
+                        key=lambda x: x['avg_response_time'] or 0,
+                        reverse=True
+                    )[:10]
+                    
+                    # 错误率最高的端点
+                    error_prone_endpoints = sorted(
+                        [ep for ep in endpoint_stats if (ep['error_rate'] or 0) > 0],
+                        key=lambda x: x['error_rate'] or 0,
+                        reverse=True
+                    )[:10]
+                    
+                    # 最热门的端点
+                    popular_endpoints = endpoint_stats[:10]
+                    
+                    return {
+                        'period': f'{start_date} to {end_date}',
+                        'all_endpoints': endpoint_stats,
+                        'slowest_endpoints': slowest_endpoints,
+                        'error_prone_endpoints': error_prone_endpoints,
+                        'popular_endpoints': popular_endpoints
+                    }
+            finally:
+                connection.close()
+                    
+        except Exception as e:
+            return {'error': str(e)}
+    
+    def get_user_behavior_analysis(self, days: int = 7) -> Dict[str, Any]:
+        """获取用户行为分析"""
+        try:
+            connection = self.pool.connection()
+            try:
+                with connection.cursor() as cursor:
+                    end_date = datetime.now().date()
+                    start_date = end_date - timedelta(days=days)
+                    
+                    # IP访问模式分析
+                    cursor.execute("""
+                        SELECT 
+                            client_ip,
+                            COUNT(*) as total_requests,
+                            COUNT(DISTINCT endpoint) as unique_endpoints,
+                            COUNT(DISTINCT DATE(timestamp)) as active_days,
+                            MIN(timestamp) as first_access,
+                            MAX(timestamp) as last_access,
+                            AVG(process_time) as avg_response_time,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as errors,
+                            SUM(CASE WHEN is_bot = 1 THEN 1 ELSE 0 END) as bot_requests,
+                            user_agent
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        GROUP BY client_ip, user_agent
+                        HAVING total_requests >= 10
+                        ORDER BY total_requests DESC
+                        LIMIT 50
+                    """, (start_date, end_date))
+                    
+                    ip_analysis = cursor.fetchall()
+                    
+                    # 设备类型统计
+                    cursor.execute("""
+                        SELECT 
+                            browser_name,
+                            os_name,
+                            COUNT(*) as request_count,
+                            COUNT(DISTINCT client_ip) as unique_users
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        AND browser_name != 'Unknown'
+                        GROUP BY browser_name, os_name
+                        ORDER BY request_count DESC
+                    """, (start_date, end_date))
+                    
+                    device_stats = cursor.fetchall()
+                    
+                    # 可疑活动检测
+                    cursor.execute("""
+                        SELECT 
+                            client_ip,
+                            COUNT(*) as requests_per_hour,
+                            COUNT(DISTINCT endpoint) as endpoints_accessed,
+                            SUM(CASE WHEN response_status = 404 THEN 1 ELSE 0 END) as not_found_errors,
+                            SUM(CASE WHEN response_status = 403 THEN 1 ELSE 0 END) as forbidden_errors,
+                            MAX(is_bot) as is_bot
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s
+                        GROUP BY client_ip
+                        HAVING requests_per_hour > 100 
+                        OR not_found_errors > 10 
+                        OR forbidden_errors > 5
+                        ORDER BY requests_per_hour DESC
+                    """, (datetime.now() - timedelta(hours=1),))
+                    
+                    suspicious_activity = cursor.fetchall()
+                    
+                    # 用户会话分析
+                    cursor.execute("""
+                        SELECT 
+                            session_id,
+                            COUNT(*) as session_requests,
+                            COUNT(DISTINCT endpoint) as endpoints_in_session,
+                            TIMESTAMPDIFF(MINUTE, MIN(timestamp), MAX(timestamp)) as session_duration,
+                            MIN(timestamp) as session_start,
+                            MAX(timestamp) as session_end
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) BETWEEN %s AND %s
+                        AND session_id IS NOT NULL
+                        GROUP BY session_id
+                        HAVING session_requests >= 5
+                        ORDER BY session_duration DESC
+                        LIMIT 20
+                    """, (start_date, end_date))
+                    
+                    session_analysis = cursor.fetchall()
+                    
+                    return {
+                        'period': f'{start_date} to {end_date}',
+                        'ip_analysis': ip_analysis,
+                        'device_statistics': device_stats,
+                        'suspicious_activity': suspicious_activity,
+                        'session_analysis': session_analysis
+                    }
+            finally:
+                connection.close()
+                    
+        except Exception as e:
+            return {'error': str(e)}
+    
+    def get_performance_alerts(self) -> Dict[str, Any]:
+        """获取性能告警信息"""
+        try:
+            connection = self.pool.connection()
+            try:
+                with connection.cursor() as cursor:
+                    now = datetime.now()
+                    last_hour = now - timedelta(hours=1)
+                    
+                    alerts = []
+                    
+                    # 检查响应时间异常
+                    cursor.execute("""
+                        SELECT endpoint, AVG(process_time) as avg_time
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s AND process_time IS NOT NULL
+                        GROUP BY endpoint
+                        HAVING avg_time > 3000
+                        ORDER BY avg_time DESC
+                    """, (last_hour,))
+                    
+                    slow_endpoints = cursor.fetchall()
+                    for endpoint in slow_endpoints:
+                        alerts.append({
+                            'type': 'SLOW_RESPONSE',
+                            'severity': 'HIGH' if (endpoint['avg_time'] or 0) > 5000 else 'MEDIUM',
+                            'message': f"端点 {endpoint['endpoint']} 平均响应时间 {endpoint['avg_time']:.0f}ms",
+                            'timestamp': now.isoformat()
+                        })
+                    
+                    # 检查错误率异常
+                    cursor.execute("""
+                        SELECT 
+                            endpoint,
+                            COUNT(*) as total,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as errors,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) / COUNT(*) * 100 as error_rate
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s
+                        GROUP BY endpoint
+                        HAVING total >= 10 AND error_rate > 10
+                        ORDER BY error_rate DESC
+                    """, (last_hour,))
+                    
+                    error_endpoints = cursor.fetchall()
+                    for endpoint in error_endpoints:
+                        alerts.append({
+                            'type': 'HIGH_ERROR_RATE',
+                            'severity': 'HIGH' if (endpoint['error_rate'] or 0) > 20 else 'MEDIUM',
+                            'message': f"端点 {endpoint['endpoint']} 错误率 {endpoint['error_rate']:.1f}%",
+                            'timestamp': now.isoformat()
+                        })
+                    
+                    # 检查异常流量
+                    cursor.execute("""
+                        SELECT 
+                            client_ip,
+                            COUNT(*) as request_count
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s
+                        GROUP BY client_ip
+                        HAVING request_count > 500
+                        ORDER BY request_count DESC
+                    """, (last_hour,))
+                    
+                    high_traffic_ips = cursor.fetchall()
+                    for ip_data in high_traffic_ips:
+                        alerts.append({
+                            'type': 'HIGH_TRAFFIC',
+                            'severity': 'MEDIUM',
+                            'message': f"IP {ip_data['client_ip']} 请求量异常: {ip_data['request_count']} 次/小时",
+                            'timestamp': now.isoformat()
+                        })
+                    
+                    # 检查系统整体负载
+                    cursor.execute("""
+                        SELECT COUNT(*) as total_requests
+                        FROM api_request_logs 
+                        WHERE timestamp >= %s
+                    """, (last_hour,))
+                    
+                    total_requests = cursor.fetchone()['total_requests']
+                    if total_requests > 10000:  # 每小时超过1万请求
+                        alerts.append({
+                            'type': 'HIGH_SYSTEM_LOAD',
+                            'severity': 'HIGH',
+                            'message': f"系统负载异常: {total_requests} 请求/小时",
+                            'timestamp': now.isoformat()
+                        })
+                    
+                    return {
+                        'alerts': alerts,
+                        'alert_count': len(alerts),
+                        'high_severity_count': len([a for a in alerts if a['severity'] == 'HIGH']),
+                        'timestamp': now.isoformat()
+                    }
+            finally:
+                connection.close()
+                    
+        except Exception as e:
+            return {'error': str(e)}
+    
+    def generate_daily_report(self, target_date: datetime = None) -> Dict[str, Any]:
+        """生成日报告"""
+        if target_date is None:
+            target_date = datetime.now().date() - timedelta(days=1)
+        
+        try:
+            connection = self.pool.connection()
+            try:
+                with connection.cursor() as cursor:
+                    # 整体统计
+                    cursor.execute("""
+                        SELECT 
+                            COUNT(*) as total_requests,
+                            COUNT(DISTINCT client_ip) as unique_ips,
+                            COUNT(DISTINCT endpoint) as unique_endpoints,
+                            AVG(process_time) as avg_response_time,
+                            MAX(process_time) as max_response_time,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) as total_errors,
+                            SUM(CASE WHEN response_status >= 400 THEN 1 ELSE 0 END) / COUNT(*) * 100 as error_rate,
+                            SUM(CASE WHEN is_bot = 1 THEN 1 ELSE 0 END) as bot_requests,
+                            SUM(CASE WHEN is_mobile = 1 THEN 1 ELSE 0 END) as mobile_requests,
+                            SUM(request_size) as total_request_size,
+                            SUM(response_size) as total_response_size
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) = %s
+                    """, (target_date,))
+                    
+                    daily_summary = cursor.fetchone()
+                    
+                    # 热门端点
+                    cursor.execute("""
+                        SELECT endpoint, COUNT(*) as requests, AVG(process_time) as avg_time
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) = %s
+                        GROUP BY endpoint
+                        ORDER BY requests DESC
+                        LIMIT 10
+                    """, (target_date,))
+                    
+                    top_endpoints = cursor.fetchall()
+                    
+                    # 错误统计
+                    cursor.execute("""
+                        SELECT response_status, COUNT(*) as count
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) = %s AND response_status >= 400
+                        GROUP BY response_status
+                        ORDER BY count DESC
+                    """, (target_date,))
+                    
+                    error_breakdown = cursor.fetchall()
+                    
+                    # 流量分布（按小时）
+                    cursor.execute("""
+                        SELECT 
+                            HOUR(timestamp) as hour,
+                            COUNT(*) as requests,
+                            AVG(process_time) as avg_time
+                        FROM api_request_logs 
+                        WHERE DATE(timestamp) = %s
+                        GROUP BY HOUR(timestamp)
+                        ORDER BY hour
+                    """, (target_date,))
+                    
+                    hourly_distribution = cursor.fetchall()
+                    
+                    return {
+                        'date': target_date.isoformat(),
+                        'summary': daily_summary,
+                        'top_endpoints': top_endpoints,
+                        'error_breakdown': error_breakdown,
+                        'hourly_distribution': hourly_distribution,
+                        'generated_at': datetime.now().isoformat()
+                    }
+            finally:
+                connection.close()
+                    
+        except Exception as e:
+            return {'error': str(e)}
+
+
+# 全局分析实例
+analytics = MonitorAnalytics()
+
+# 导出分析函数
+def get_realtime_metrics():
+    """获取实时监控指标"""
+    return analytics.get_realtime_metrics()
+
+def get_traffic_trend(days: int = 7):
+    """获取流量趋势"""
+    return analytics.get_traffic_trend(days)
+
+def get_endpoint_analysis(days: int = 7):
+    """获取端点分析"""
+    return analytics.get_endpoint_analysis(days)
+
+def get_user_behavior_analysis(days: int = 7):
+    """获取用户行为分析"""
+    return analytics.get_user_behavior_analysis(days)
+
+def get_performance_alerts():
+    """获取性能告警"""
+    return analytics.get_performance_alerts()
+
+def generate_daily_report(target_date: datetime = None):
+    """生成日报告"""
+    return analytics.generate_daily_report(target_date)