workerman=一个完整的进程健康监控系统。
小知识：
Workerman 不同进程之间内存隔离，无法直接共享变量
health_monitor.php
<?php
/**
* Workerman 子进程健康监控系统
* 功能：心跳检测、状态监控、异常告警、自动恢复
*/
require_once __DIR__ . '/vendor/autoload.php';
use Workerman\Worker;
use Workerman\Timer;
use Workerman\Connection\TcpConnection;
// ============================================
// 全局变量：存储健康状态数据
// ============================================
class HealthMonitor {
// 存储每个进程的健康数据
public static $healthData = [];
// 心跳超时时间（秒）
public static $heartbeatTimeout = 10;
// 健康检查间隔（秒）
public static $checkInterval = 5;
// CPU 告警阈值（%）
public static $cpuThreshold = 80;
// 内存告警阈值（%）
public static $memoryThreshold = 80;
// 异常重启次数阈值
public static $maxRestarts = 5;
// 重启计数器
public static $restartCount = [];
}
// ============================================
// 业务 Worker：被监控的工作进程
// ============================================
$business_worker = new Worker('tcp://0.0.0.0:8888');
$business_worker->name = 'BusinessWorker';
$business_worker->count = 4;
$business_worker->onWorkerStart = function($worker) {
$pid = posix_getpid();
$workerId = $worker->id;
echo "[BusinessWorker 启动] PID: {$pid}, Worker ID: {$workerId}\n";
// 初始化健康数据
HealthMonitor::$healthData[$pid] = [
'pid' => $pid,
'worker_id' => $workerId,
'worker_name' => $worker->name,
'status' => 'healthy',
'last_heartbeat' => time(),
'start_time' => time(),
'cpu' => 0,
'memory' => 0,
'memory_mb' => 0,
'connections' => 0,
'requests' => 0,
'errors' => 0,
'warnings' => []
];
// 每 3 秒发送心跳
Timer::add(3, function() use ($pid, $workerId, $worker) {
// 更新心跳时间
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['last_heartbeat'] = time();
HealthMonitor::$healthData[$pid]['connections'] = count($worker->connections);
// 获取进程资源使用情况
$stats = getProcessStats($pid);
if ($stats) {
HealthMonitor::$healthData[$pid]['cpu'] = $stats['cpu'];
HealthMonitor::$healthData[$pid]['memory'] = $stats['memory'];
HealthMonitor::$healthData[$pid]['memory_mb'] = $stats['memoryMB'];
// 检查告警
checkHealthWarnings($pid);
}
}
});
// 模拟处理任务
Timer::add(5, function() use ($pid, $workerId) {
echo "[{$pid}] Worker {$workerId} 处理任务...\n";
// 模拟随机错误
if (rand(1, 20) == 1) {
echo "[{$pid}] Worker {$workerId} 发生错误！\n";
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['errors']++;
}
}
// 更新请求计数
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['requests']++;
}
});
};
$business_worker->onConnect = function($connection) {
$pid = posix_getpid();
echo "[{$pid}] 新连接\n";
$connection->send("连接成功！由进程 {$pid} 为您服务\n");
};
$business_worker->onMessage = function($connection, $data) {
$pid = posix_getpid();
// 更新请求计数
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['requests']++;
}
$connection->send("进程 {$pid} 已处理: {$data}");
};
$business_worker->onWorkerStop = function($worker) {
$pid = posix_getpid();
echo "[BusinessWorker 停止] PID: {$pid}\n";
// 清理健康数据
if (isset(HealthMonitor::$healthData[$pid])) {
unset(HealthMonitor::$healthData[$pid]);
}
};
// ============================================
// 监控 Worker：监控所有进程健康状态
// ============================================
$monitor_worker = new Worker();
$monitor_worker->name = 'HealthMonitor';
$monitor_worker->count = 1;
$monitor_worker->onWorkerStart = function($worker) {
echo "[HealthMonitor 启动] 健康监控服务已启动\n";
echo "检查间隔: " . HealthMonitor::$checkInterval . " 秒\n";
echo "心跳超时: " . HealthMonitor::$heartbeatTimeout . " 秒\n\n";
// 定期健康检查
Timer::add(HealthMonitor::$checkInterval, function() {
performHealthCheck();
});
// 每 10 秒输出健康报告
Timer::add(10, function() {
printHealthReport();
});
};
// ============================================
// WebSocket 监控服务：实时推送监控数据
// ============================================
$ws_worker = new Worker("websocket://0.0.0.0:8282");
$ws_worker->name = 'MonitorWebSocket';
$ws_worker->count = 1;
$ws_worker->onConnect = function($connection) {
echo "[WebSocket] 新客户端连接\n";
};
$ws_worker->onMessage = function($connection, $data) {
$message = json_decode($data, true);
if (isset($message['action']) && $message['action'] === 'getHealth') {
$healthData = prepareHealthData();
$connection->send(json_encode([
'action' => 'healthUpdate',
'data' => $healthData,
'timestamp' => date('Y-m-d H:i:s')
]));
}
};
$ws_worker->onWorkerStart = function($worker) {
// 每 2 秒推送健康数据给所有客户端
Timer::add(2, function() use ($worker) {
$healthData = prepareHealthData();
$data = json_encode([
'action' => 'healthUpdate',
'data' => $healthData,
'timestamp' => date('Y-m-d H:i:s')
]);
foreach($worker->connections as $connection) {
$connection->send($data);
}
});
};
// ============================================
// HTTP 服务：提供监控页面
// ============================================
$http_worker = new Worker("http://0.0.0.0:8080");
$http_worker->name = 'HttpServer';
$http_worker->count = 1;
$http_worker->onMessage = function($connection, $request) {
$path = $request->path();
if ($path === '/' || $path === '/health.html') {
$htmlFile = __DIR__ . '/health_monitor.html';
if (file_exists($htmlFile)) {
$connection->send(file_get_contents($htmlFile));
} else {
$connection->send("HTTP/1.1 404 Not Found\r\n\r\nhealth_monitor.html not found");
}
} elseif ($path === '/api/health') {
$healthData = prepareHealthData();
$connection->send(json_encode($healthData));
} else {
$connection->send("HTTP/1.1 404 Not Found\r\n\r\n404 Not Found");
}
};
// ============================================
// 辅助函数
// ============================================
/**
* 获取进程统计信息
*/
function getProcessStats($pid) {
if (!posix_kill($pid, 0)) {
return null;
}
$cmd = sprintf("ps -p %d -o %%cpu,%%mem,rss --no-headers 2>/dev/null", $pid);
$output = shell_exec($cmd);
if (empty($output)) {
return null;
}
$parts = preg_split('/\s+/', trim($output));
if (count($parts) < 3) {
return null;
}
return [
'cpu' => (float)$parts[0],
'memory' => (float)$parts[1],
'memoryMB' => round((float)$parts[2] / 1024, 2)
];
}
/**
* 检查健康告警
*/
function checkHealthWarnings($pid) {
if (!isset(HealthMonitor::$healthData[$pid])) {
return;
}
$data = &HealthMonitor::$healthData[$pid];
$data['warnings'] = [];
$data['status'] = 'healthy';
// 检查 CPU
if ($data['cpu'] > HealthMonitor::$cpuThreshold) {
$data['warnings'][] = "CPU 使用率过高: {$data['cpu']}%";
$data['status'] = 'warning';
echo "[告警] PID {$pid} CPU 使用率过高: {$data['cpu']}%\n";
}
// 检查内存
if ($data['memory'] > HealthMonitor::$memoryThreshold) {
$data['warnings'][] = "内存使用率过高: {$data['memory']}%";
$data['status'] = 'warning';
echo "[告警] PID {$pid} 内存使用率过高: {$data['memory']}%\n";
}
// 检查错误率
if ($data['requests'] > 0) {
$errorRate = ($data['errors'] / $data['requests']) * 100;
if ($errorRate > 10) {
$data['warnings'][] = sprintf("错误率过高: %.2f%%", $errorRate);
$data['status'] = 'warning';
echo "[告警] PID {$pid} 错误率过高: {$errorRate}%\n";
}
}
}
/**
* 执行健康检查
*/
function performHealthCheck() {
$now = time();
$unhealthyCount = 0;
foreach (HealthMonitor::$healthData as $pid => &$data) {
// 检查心跳超时
$timeSinceHeartbeat = $now - $data['last_heartbeat'];
if ($timeSinceHeartbeat > HealthMonitor::$heartbeatTimeout) {
$data['status'] = 'unhealthy';
$data['warnings'][] = "心跳超时: {$timeSinceHeartbeat} 秒";
echo "[严重] PID {$pid} 心跳超时！已超时 {$timeSinceHeartbeat} 秒\n";
$unhealthyCount++;
// 检查进程是否真的死了
if (!posix_kill($pid, 0)) {
echo "[严重] PID {$pid} 进程已死亡！\n";
unset(HealthMonitor::$healthData[$pid]);
// 记录重启次数
if (!isset(HealthMonitor::$restartCount[$pid])) {
HealthMonitor::$restartCount[$pid] = 0;
}
HealthMonitor::$restartCount[$pid]++;
// 检查是否超过重启次数限制
if (HealthMonitor::$restartCount[$pid] > HealthMonitor::$maxRestarts) {
echo "[严重] PID {$pid} 重启次数过多，需要人工介入！\n";
}
}
}
}
if ($unhealthyCount > 0) {
echo "[健康检查] 发现 {$unhealthyCount} 个不健康的进程\n";
}
}
/**
* 打印健康报告
*/
function printHealthReport() {
echo "\n========================================\n";
echo "健康监控报告 - " . date('Y-m-d H:i:s') . "\n";
echo "========================================\n";
$totalProcesses = count(HealthMonitor::$healthData);
$healthyCount = 0;
$warningCount = 0;
$unhealthyCount = 0;
foreach (HealthMonitor::$healthData as $data) {
switch ($data['status']) {
case 'healthy':
$healthyCount++;
break;
case 'warning':
$warningCount++;
break;
case 'unhealthy':
$unhealthyCount++;
break;
}
}
echo "总进程数: {$totalProcesses}\n";
echo "健康: {$healthyCount} | 警告: {$warningCount} | 异常: {$unhealthyCount}\n";
echo "========================================\n";
foreach (HealthMonitor::$healthData as $data) {
$uptime = time() - $data['start_time'];
$statusSymbol = $data['status'] === 'healthy' ? '✓' : ($data['status'] === 'warning' ? '⚠' : '✗');
echo "{$statusSymbol} PID {$data['pid']} (Worker {$data['worker_id']})\n";
echo " 状态: {$data['status']}\n";
echo " CPU: {$data['cpu']}% | 内存: {$data['memory']}% ({$data['memory_mb']}MB)\n";
echo " 连接数: {$data['connections']} | 请求数: {$data['requests']} | 错误数: {$data['errors']}\n";
echo " 运行时长: " . formatUptime($uptime) . "\n";
if (!empty($data['warnings'])) {
echo " 告警: " . implode(', ', $data['warnings']) . "\n";
}
echo "\n";
}
}
/**
* 准备健康数据用于 API/WebSocket
* 直接从系统获取所有进程信息
*/
function prepareHealthData() {
$processes = [];
// 查找主进程（多种方式尝试）
$masterPid = 0;
// 方式1: 通过 master process 标识查找
$cmd = "ps -ef | grep 'master process' | grep 'health_monitor.php' | grep -v grep | awk '{print $2}' | head -1";
$masterPid = (int)trim(shell_exec($cmd));
// 方式2: 如果方式1失败，通过 Workerman 标识查找
if (!$masterPid) {
$cmd = "ps -ef | grep 'Workerman\\[health_monitor.php\\]' | grep -v grep | awk '{print $2}' | head -1";
$masterPid = (int)trim(shell_exec($cmd));
}
// 方式3: 如果还是失败，查找 health_monitor.php 的父进程
if (!$masterPid) {
$cmd = "ps -ef | grep 'health_monitor.php' | grep -v grep | awk '{print $2}' | head -1";
$masterPid = (int)trim(shell_exec($cmd));
}
if (!$masterPid) {
return [
'summary' => [
'total' => 0,
'healthy' => 0,
'warning' => 0,
'unhealthy' => 0
],
'processes' => []
];
}
// 获取所有子进程
$cmd = "ps --ppid {$masterPid} -o pid,cmd --no-headers 2>/dev/null";
$output = shell_exec($cmd);
if (empty($output)) {
return [
'summary' => [
'total' => 0,
'healthy' => 0,
'warning' => 0,
'unhealthy' => 0
],
'processes' => []
];
}
$lines = explode("\n", trim($output));
$healthyCount = 0;
$warningCount = 0;
$unhealthyCount = 0;
foreach ($lines as $line) {
if (empty($line)) continue;
$parts = preg_split('/\s+/', trim($line), 2);
$pid = (int)$parts[0];
$cmd = $parts[1] ?? '';
// 只监控 BusinessWorker
if (strpos($cmd, 'BusinessWorker') === false) {
continue;
}
// 获取进程统计信息
$stats = getProcessStats($pid);
if (!$stats) {
continue;
}
// 确定 Worker ID
preg_match('/worker process BusinessWorker (\d+)/', $cmd, $matches);
$workerId = isset($matches[1]) ? (int)$matches[1] : 0;
// 判断健康状态
$status = 'healthy';
$warnings = [];
if ($stats['cpu'] > HealthMonitor::$cpuThreshold) {
$status = 'warning';
$warnings[] = "CPU 使用率过高: {$stats['cpu']}%";
}
if ($stats['memory'] > HealthMonitor::$memoryThreshold) {
$status = 'warning';
$warnings[] = "内存使用率过高: {$stats['memory']}%";
}
// 统计状态
switch ($status) {
case 'healthy': $healthyCount++; break;
case 'warning': $warningCount++; break;
case 'unhealthy': $unhealthyCount++; break;
}
// 获取进程启动时间
$etimeCmd = "ps -p {$pid} -o etime --no-headers 2>/dev/null";
$etime = trim(shell_exec($etimeCmd));
$startTime = time() - parseElapsedTime($etime);
$processes[] = [
'pid' => $pid,
'worker_id' => $workerId,
'worker_name' => 'BusinessWorker',
'status' => $status,
'last_heartbeat' => time(),
'start_time' => $startTime,
'cpu' => $stats['cpu'],
'memory' => $stats['memory'],
'memory_mb' => $stats['memoryMB'],
'connections' => 0, // 无法跨进程获取
'requests' => 0, // 无法跨进程获取
'errors' => 0, // 无法跨进程获取
'warnings' => $warnings
];
}
return [
'summary' => [
'total' => count($processes),
'healthy' => $healthyCount,
'warning' => $warningCount,
'unhealthy' => $unhealthyCount
],
'processes' => $processes
];
}
/**
* 解析 etime 格式转换为秒数
*/
function parseElapsedTime($etime) {
$etime = trim($etime);
$seconds = 0;
// 格式: [[dd-]hh:]mm:ss
if (strpos($etime, '-') !== false) {
list($days, $time) = explode('-', $etime);
$seconds += (int)$days * 86400;
} else {
$time = $etime;
}
$parts = array_reverse(explode(':', $time));
$multipliers = [1, 60, 3600]; // 秒、分、时
foreach ($parts as $i => $value) {
$seconds += (int)$value * $multipliers[$i];
}
return $seconds;
}
/**
* 格式化运行时长
*/
function formatUptime($seconds) {
$hours = floor($seconds / 3600);
$minutes = floor(($seconds % 3600) / 60);
$secs = $seconds % 60;
return sprintf("%02d:%02d:%02d", $hours, $minutes, $secs);
}
// ============================================
// 启动信息
// ============================================
echo "======================================\n";
echo "Workerman 健康监控系统\n";
echo "======================================\n";
echo "BusinessWorker : tcp://0.0.0.0:8888 (4个进程)\n";
echo "HealthMonitor : 健康检查服务\n";
echo "WebSocket : ws://0.0.0.0:8282\n";
echo "监控页面 : http://localhost:8080/health.html\n";
echo "======================================\n\n";
// 运行所有 Worker
Worker::runAll();
解释：
这是一个完整的 Workerman 子进程健康监控系统，用于监控所有 Worker 进程的健康状态。
📋 第 1-11 行：文件头和依赖导入
<?php
/**
- Workerman 子进程健康监控系统
-
功能：心跳检测、状态监控、异常告警、自动恢复
*/require_once __DIR__ . '/vendor/autoload.php';
use Workerman\Worker;
use Workerman\Timer;
use Workerman\Connection\TcpConnection;作用：
- 加载 Composer 自动加载器
- 导入必要的 Workerman 类
- Worker - 核心进程类
- Timer - 定时器
- TcpConnection - TCP 连接类
🔧 第 16-37 行：配置类 HealthMonitor
class HealthMonitor {
// 存储每个进程的健康数据
public static $healthData = [];// 心跳超时时间（秒）
public static $heartbeatTimeout = 10;// 健康检查间隔（秒）
public static $checkInterval = 5;// CPU 告警阈值（%）
public static $cpuThreshold = 80;// 内存告警阈值（%）
public static $memoryThreshold = 80;// 异常重启次数阈值
public static $maxRestarts = 5;// 重启计数器
public static $restartCount = [];
}详细解释：
配置项 默认值 说明 $healthData [] 存储所有进程的健康数据（进程内存隔离，只在当前进程有效） $heartbeatTimeout 10 秒 心跳超时阈值，超过此时间未收到心跳视为异常 $checkInterval 5 秒 健康检查的执行间隔 $cpuThreshold 80% CPU 使用率告警阈值，超过触发警告 $memoryThreshold 80% 内存使用率告警阈值 $maxRestarts 5 次 进程最大重启次数，超过需人工介入 $restartCount [] 记录每个进程的重启次数 为什么用 static？
- 静态属性在类级别共享
- 可以在不同函数间访问同一份数据
- 注意：多进程间仍然隔离
🟦 第 42-133 行：BusinessWorker - 业务进程
创建 Worker（第 42-44 行）
$business_worker = new Worker('tcp://0.0.0.0:8888');
$business_worker->name = 'BusinessWorker';
$business_worker->count = 4;作用：
- 创建 TCP 服务器，监听 8888 端口
- 命名为 BusinessWorker
- 创建 4 个子进程（充分利用多核 CPU）
onWorkerStart - 进程启动（第 46-106 行）
$business_worker->onWorkerStart = function($worker) {
$pid = posix_getpid();
$workerId = $worker->id;echo "[BusinessWorker 启动] PID: {$pid}, Worker ID: {$workerId}\n";
触发时机： 每个子进程启动时执行一次
- 初始化健康数据（第 52-67 行）
HealthMonitor::$healthData[$pid] = [
'pid' => $pid, // 进程 ID
'worker_id' => $workerId, // Worker ID (0-3)
'worker_name' => $worker->name, // Worker 名称
'status' => 'healthy', // 健康状态
'last_heartbeat' => time(), // 最后心跳时间
'start_time' => time(), // 启动时间
'cpu' => 0, // CPU 使用率
'memory' => 0, // 内存使用率
'memory_mb' => 0, // 内存占用 MB
'connections' => 0, // 连接数
'requests' => 0, // 请求数
'errors' => 0, // 错误数
'warnings' => [] // 告警列表
];作用： 为当前进程创建健康数据结构
注意：
- 这个数据只在当前子进程内有效
- 由于进程隔离，其他进程看不到这份数据
- 这就是为什么后面需要用 ps 命令跨进程获取数据
- 心跳定时器（第 69-87 行）
Timer::add(3, function() use ($pid, $workerId, $worker) {
// 更新心跳时间
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['last_heartbeat'] = time();
HealthMonitor::$healthData[$pid]['connections'] = count($worker->connections);// 获取进程资源使用情况 $stats = getProcessStats($pid); if ($stats) { HealthMonitor::$healthData[$pid]['cpu'] = $stats['cpu']; HealthMonitor::$healthData[$pid]['memory'] = $stats['memory']; HealthMonitor::$healthData[$pid]['memory_mb'] = $stats['memoryMB']; // 检查告警 checkHealthWarnings($pid); }}
});执行周期： 每 3 秒
工作流程：
- 更新心跳时间戳 last_heartbeat
- 统计当前连接数
- 调用 getProcessStats() 获取 CPU/内存使用情况
- 调用 checkHealthWarnings() 检查是否需要告警
心跳机制：
子进程 监控进程
│ │
├─ 每3秒更新 last_heartbeat │
│ │
│ 每5秒检查
│ 心跳是否超时
│ │
│ 如果超过10秒未更新 ──────────> 触发告警
- 模拟任务定时器（第 89-105 行）
Timer::add(5, function() use ($pid, $workerId) {
echo "[{$pid}] Worker {$workerId} 处理任务...\n";// 模拟随机错误（5% 概率）
if (rand(1, 20) == 1) {
echo "[{$pid}] Worker {$workerId} 发生错误！\n";
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['errors']++;
}
}// 更新请求计数
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['requests']++;
}
});执行周期： 每 5 秒
作用：
- 模拟业务处理
- 随机产生错误（5% 概率，用于测试错误率告警）
- 更新请求计数器
onConnect - 连接回调（第 108-112 行）
$business_worker->onConnect = function($connection) {
$pid = posix_getpid();
echo "[{$pid}] 新连接\n";
$connection->send("连接成功！由进程 {$pid} 为您服务\n");
};触发时机： 客户端连接到 8888 端口时
onMessage - 消息回调（第 114-123 行）
$business_worker->onMessage = function($connection, $data) {
$pid = posix_getpid();// 更新请求计数
if (isset(HealthMonitor::$healthData[$pid])) {
HealthMonitor::$healthData[$pid]['requests']++;
}$connection->send("进程 {$pid} 已处理: {$data}");
};触发时机： 客户端发送数据时
作用： 统计请求数
onWorkerStop - 停止回调（第 125-133 行）
$business_worker->onWorkerStop = function($worker) {
$pid = posix_getpid();
echo "[BusinessWorker 停止] PID: {$pid}\n";// 清理健康数据
if (isset(HealthMonitor::$healthData[$pid])) {
unset(HealthMonitor::$healthData[$pid]);
}
};触发时机： 进程退出前
作用： 清理资源
🟩 第 138-156 行：HealthMonitor - 监控进程
$monitor_worker = new Worker();
$monitor_worker->name = 'HealthMonitor';
$monitor_worker->count = 1;$monitor_worker->onWorkerStart = function($worker) {
echo "[HealthMonitor 启动] 健康监控服务已启动\n";
echo "检查间隔: " . HealthMonitor::$checkInterval . " 秒\n";
echo "心跳超时: " . HealthMonitor::$heartbeatTimeout . " 秒\n\n";// 定期健康检查
Timer::add(HealthMonitor::$checkInterval, function() {
performHealthCheck(); // 每 5 秒执行健康检查
});// 每 10 秒输出健康报告
Timer::add(10, function() {
printHealthReport(); // 每 10 秒打印报告
});
};关键点：
- new Worker() - 没有监听地址，纯后台监控进程
- count = 1 - 只需要 1 个监控进程
- 两个定时器：
- 5 秒一次健康检查
- 10 秒一次输出报告
🟨 第 161-196 行：WebSocket - 实时推送
$ws_worker = new Worker("websocket://0.0.0.0:8282");
$ws_worker->name = 'MonitorWebSocket';
$ws_worker->count = 1;onMessage - 处理客户端请求（第 169-180 行）
$ws_worker->onMessage = function($connection, $data) {
$message = json_decode($data, true);if (isset($message['action']) && $message['action'] === 'getHealth') {
$healthData = prepareHealthData();
$connection->send(json_encode([
'action' => 'healthUpdate',
'data' => $healthData,
'timestamp' => date('Y-m-d H:i:s')
]));
}
};作用： 当客户端请求健康数据时，立即返回
onWorkerStart - 自动推送（第 182-196 行）
$ws_worker->onWorkerStart = function($worker) {
// 每 2 秒推送健康数据给所有客户端
Timer::add(2, function() use ($worker) {
$healthData = prepareHealthData();
$data = json_encode([
'action' => 'healthUpdate',
'data' => $healthData,
'timestamp' => date('Y-m-d H:i:s')
]);foreach($worker->connections as $connection) { $connection->send($data); }});
};作用： 每 2 秒自动推送数据给所有连接的浏览器
🟪 第 201-221 行：HTTP Server - 监控页面
$http_worker = new Worker("http://0.0.0.0:8080");
$http_worker->name = 'HttpServer';
$http_worker->count = 1;$http_worker->onMessage = function($connection, $request) {
$path = $request->path();if ($path === '/' || $path === '/health.html') {
$htmlFile = __DIR__ . '/health_monitor.html';
if (file_exists($htmlFile)) {
$connection->send(file_get_contents($htmlFile));
} else {
$connection->send("HTTP/1.1 404 Not Found\r\n\r\nhealth_monitor.html not found");
}
} elseif ($path === '/api/health') {
$healthData = prepareHealthData();
$connection->send(json_encode($healthData));
} else {
$connection->send("HTTP/1.1 404 Not Found\r\n\r\n404 Not Found");
}
};路由：
- / 或 /health.html - 返回监控页面HTML
- /api/health - 返回JSON格式的健康数据
- 其他 - 404
🔧 辅助函数详解
getProcessStats() - 获取进程统计（第 230-253 行）
function getProcessStats($pid) {
if (!posix_kill($pid, 0)) {
return null;
}$cmd = sprintf("ps -p %d -o %%cpu,%%mem,rss --no-headers 2>/dev/null", $pid);
$output = shell_exec($cmd);if (empty($output)) {
return null;
}$parts = preg_split('/\s+/', trim($output));
if (count($parts) < 3) {
return null;
}return [
'cpu' => (float)$parts[0],
'memory' => (float)$parts[1],
'memoryMB' => round((float)$parts[2] / 1024, 2)
];
}工作流程：
- posix_kill($pid, 0) - 检查进程是否存在（不杀死，只检查）
- 执行 ps 命令获取 CPU、内存、RSS
- 解析输出并返回数组
返回示例：
[
'cpu' => 12.5, // CPU 使用率 12.5%
'memory' => 2.3, // 内存使用率 2.3%
'memoryMB' => 45.6 // 实际内存占用 45.6MB
]
checkHealthWarnings() - 检查告警（第 258-290 行）
function checkHealthWarnings($pid) {
if (!isset(HealthMonitor::$healthData[$pid])) {
return;
}$data = &HealthMonitor::$healthData[$pid];
$data['warnings'] = [];
$data['status'] = 'healthy';// 检查 CPU
if ($data['cpu'] > HealthMonitor::$cpuThreshold) {
$data['warnings'][] = "CPU 使用率过高: {$data['cpu']}%";
$data['status'] = 'warning';
echo "[告警] PID {$pid} CPU 使用率过高: {$data['cpu']}%\n";
}// 检查内存
if ($data['memory'] > HealthMonitor::$memoryThreshold) {
$data['warnings'][] = "内存使用率过高: {$data['memory']}%";
$data['status'] = 'warning';
echo "[告警] PID {$pid} 内存使用率过高: {$data['memory']}%\n";
}// 检查错误率
if ($data['requests'] > 0) {
$errorRate = ($data['errors'] / $data['requests']) * 100;
if ($errorRate > 10) {
$data['warnings'][] = sprintf("错误率过高: %.2f%%", $errorRate);
$data['status'] = 'warning';
echo "[告警] PID {$pid} 错误率过高: {$errorRate}%\n";
}
}
}告警条件：
- CPU > 80%
- 内存 > 80%
- 错误率 > 10%
状态变化：
- 正常：status = 'healthy'
- 触发告警：status = 'warning'
performHealthCheck() - 健康检查（第 295-331 行）
function performHealthCheck() {
$now = time();
$unhealthyCount = 0;foreach (HealthMonitor::$healthData as $pid => &$data) {
// 检查心跳超时
$timeSinceHeartbeat = $now - $data['last_heartbeat'];if ($timeSinceHeartbeat > HealthMonitor::$heartbeatTimeout) { $data['status'] = 'unhealthy'; $data['warnings'][] = "心跳超时: {$timeSinceHeartbeat} 秒"; echo "[严重] PID {$pid} 心跳超时！已超时 {$timeSinceHeartbeat} 秒\n"; $unhealthyCount++; // 检查进程是否真的死了 if (!posix_kill($pid, 0)) { echo "[严重] PID {$pid} 进程已死亡！\n"; unset(HealthMonitor::$healthData[$pid]); // 记录重启次数 if (!isset(HealthMonitor::$restartCount[$pid])) { HealthMonitor::$restartCount[$pid] = 0; } HealthMonitor::$restartCount[$pid]++; // 检查是否超过重启次数限制 if (HealthMonitor::$restartCount[$pid] > HealthMonitor::$maxRestarts) { echo "[严重] PID {$pid} 重启次数过多，需要人工介入！\n"; } } }}
if ($unhealthyCount > 0) {
echo "[健康检查] 发现 {$unhealthyCount} 个不健康的进程\n";
}
}检查流程：
- 计算距离上次心跳的时间
- 如果超过 10 秒 → 标记为 unhealthy
- 使用 posix_kill($pid, 0) 检查进程是否真的死了
- 如果死了：
- 从健康数据中移除
- 记录重启次数
- 超过 5 次重启 → 需要人工介入
prepareHealthData() - 准备数据（第 385-515 行）
这是最关键的函数，用于跨进程获取数据。
function prepareHealthData() {
$processes = [];// 查找主进程（多种方式尝试）
$masterPid = 0;// 方式1: 通过 master process 标识查找
$cmd = "ps -ef | grep 'master process' | grep 'health_monitor.php' | grep -v grep | awk '{print $2}' | head -1";
$masterPid = (int)trim(shell_exec($cmd));// 方式2: 如果方式1失败，通过 Workerman 标识查找
if (!$masterPid) {
$cmd = "ps -ef | grep 'Workerman\[health_monitor.php\]' | grep -v grep | awk '{print $2}' | head -1";
$masterPid = (int)trim(shell_exec($cmd));
}// 方式3: 如果还是失败，查找 health_monitor.php 的父进程
if (!$masterPid) {
$cmd = "ps -ef | grep 'health_monitor.php' | grep -v grep | awk '{print $2}' | head -1";
$masterPid = (int)trim(shell_exec($cmd));
}为什么需要多种方式？
- 不同系统下进程名格式不同
- WSL、Linux、macOS 的 ps 输出有差异
- 多种fallback确保能找到主进程
// 获取所有子进程
$cmd = "ps --ppid {$masterPid} -o pid,cmd --no-headers 2>/dev/null";
$output = shell_exec($cmd);ps --ppid 命令：
- --ppid {$masterPid} - 查找父进程为主进程的所有子进程
- -o pid,cmd - 只输出 PID 和命令
- --no-headers - 不显示表头
foreach ($lines as $line) {
if (empty($line)) continue;$parts = preg_split('/\s+/', trim($line), 2); $pid = (int)$parts[0]; $cmd = $parts[1] ?? ''; // 只监控 BusinessWorker if (strpos($cmd, 'BusinessWorker') === false) { continue; }过滤逻辑：
- 只监控包含 "BusinessWorker" 的进程
- 跳过 HealthMonitor、WebSocket、HttpServer
最终返回格式：
return [
'summary' => [
'total' => 4,
'healthy' => 3,
'warning' => 1,
'unhealthy' => 0
],
'processes' => [
[
'pid' => 1639,
'worker_id' => 0,
'status' => 'healthy',
'cpu' => 12.5,
'memory' => 2.3,
...
],
...
]
];
🚀 第 556-566 行：启动
echo "======================================\n";
echo "Workerman 健康监控系统\n";
echo "======================================\n";
echo "BusinessWorker : tcp://0.0.0.0:8888 (4个进程)\n";
echo "HealthMonitor : 健康检查服务\n";
echo "WebSocket : ws://0.0.0.0:8282\n";
echo "监控页面 : http://localhost:8080/health.html\n";
echo "======================================\n\n";// 运行所有 Worker
Worker::runAll();Worker::runAll() 做了什么？
- Fork 所有子进程（4+1+1+1 = 7个）
- 初始化每个子进程
- 进入事件循环
- 阻塞运行，直到收到停止信号
📊 系统架构总结
主进程 (Master)
├── BusinessWorker #0 (PID: 1639) ──┐
├── BusinessWorker #1 (PID: 1640) ├─ 被监控的业务进程
├── BusinessWorker #2 (PID: 1641) │ 每3秒发送心跳
├── BusinessWorker #3 (PID: 1642) ──┘
│
├── HealthMonitor (PID: xxx) ────────── 监控进程
│ ├── 每5秒健康检查 检查心跳超时
│ └── 每10秒输出报告 检查资源使用
│
├── WebSocket (PID: xxx) ────────────── WebSocket 服务
│ └── 每2秒推送数据到浏览器 实时更新
│
└── HttpServer (PID: xxx) ───────────── HTTP 服务
├── GET / → 监控页面
└── GET /api/health → JSON 数据
这就是完整的健康监控系统！核心思想是：
- 业务进程发送心跳
- 监控进程检查心跳和资源
- WebSocket 推送数据到浏览器
- HTTP Server 提供监控页面
所有进程协同工作，实现完整的健康监控！🎉
health_monitor.html
<!DOCTYPE html> <html lang="zh-CN"> <head> <meta charset="UTF-8"> <meta name="viewport" content="width=device-width, initial-scale=1.0"> <title>进程健康监控</title> <style> * { margin: 0; padding: 0; box-sizing: border-box; } body { font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif; background: linear-gradient(135deg, #1e3c72 0%, #2a5298 100%); padding: 20px; color: #333; } .container { max-width: 1400px; margin: 0 auto; } .header { background: white; border-radius: 10px; padding: 20px; margin-bottom: 20px; box-shadow: 0 4px 6px rgba(0,0,0,0.1); } .header h1 { color: #1e3c72; margin-bottom: 10px; display: flex; align-items: center; gap: 10px; } .status-bar { display: flex; gap: 20px; align-items: center; margin-top: 10px; } .status-indicator { display: flex; align-items: center; gap: 8px; } .status-dot { width: 12px; height: 12px; border-radius: 50%; background: #22c55e; animation: pulse 2s infinite; } .status-dot.disconnected { background: #ef4444; animation: none; } @keyframes pulse { 0%, 100% { opacity: 1; } 50% { opacity: 0.5; } } .summary-cards { display: grid; grid-template-columns: repeat(auto-fit, minmax(250px, 1fr)); gap: 20px; margin-bottom: 20px; } .card { background: white; border-radius: 10px; padding: 20px; box-shadow: 0 4px 6px rgba(0,0,0,0.1); transition: transform 0.2s; } .card:hover { transform: translateY(-5px); } .card h3 { color: #666; font-size: 14px; margin-bottom: 10px; text-transform: uppercase; } .card .value { font-size: 36px; font-weight: bold; margin-bottom: 5px; } .card.healthy .value { color: #22c55e; } .card.warning .value { color: #f59e0b; } .card.unhealthy .value { color: #ef4444; } .processes-section { background: white; border-radius: 10px; padding: 20px; box-shadow: 0 4px 6px rgba(0,0,0,0.1); } .processes-section h2 { color: #1e3c72; margin-bottom: 20px; } .process-item { background: #f8f9fa; border-radius: 8px; padding: 15px; margin-bottom: 15px; border-left: 4px solid #22c55e; transition: all 0.3s; } .process-item.warning { border-left-color: #f59e0b; background: #fffbeb; } .process-item.unhealthy { border-left-color: #ef4444; background: #fef2f2; animation: shake 0.5s; } @keyframes shake { 0%, 100% { transform: translateX(0); } 25% { transform: translateX(-10px); } 75% { transform: translateX(10px); } } .process-header { display: flex; justify-content: space-between; align-items: center; margin-bottom: 10px; } .process-title { font-size: 18px; font-weight: bold; display: flex; align-items: center; gap: 10px; } .status-badge { display: inline-block; padding: 4px 12px; border-radius: 20px; font-size: 12px; font-weight: bold; } .status-badge.healthy { background: #dcfce7; color: #16a34a; } .status-badge.warning { background: #fef3c7; color: #d97706; } .status-badge.unhealthy { background: #fee2e2; color: #dc2626; } .process-stats { display: grid; grid-template-columns: repeat(auto-fit, minmax(150px, 1fr)); gap: 15px; margin-top: 15px; } .stat-item { background: white; padding: 10px; border-radius: 5px; } .stat-label { font-size: 12px; color: #666; margin-bottom: 5px; } .stat-value { font-size: 20px; font-weight: bold; color: #333; } .progress-bar { width: 100%; height: 20px; background: #e5e7eb; border-radius: 10px; overflow: hidden; margin-top: 5px; } .progress-fill { height: 100%; background: linear-gradient(90deg, #22c55e 0%, #16a34a 100%); transition: width 0.3s ease; display: flex; align-items: center; justify-content: center; color: white; font-size: 12px; font-weight: bold; } .progress-fill.warning { background: linear-gradient(90deg, #f59e0b 0%, #d97706 100%); } .progress-fill.danger { background: linear-gradient(90deg, #ef4444 0%, #dc2626 100%); } .warnings { background: #fef3c7; border-left: 3px solid #f59e0b; padding: 10px; border-radius: 5px; margin-top: 10px; } .warnings h4 { color: #d97706; margin-bottom: 5px; font-size: 14px; } .warnings ul { list-style: none; padding-left: 0; } .warnings li { color: #d97706; padding: 3px 0; font-size: 13px; } .warnings li:before { content: "⚠ "; margin-right: 5px; } .no-data { text-align: center; padding: 40px; color: #999; } .last-update { text-align: center; color: white; margin-top: 20px; font-size: 14px; } .status-icon { font-size: 24px; } .stat-row { display: flex; gap: 10px; margin-bottom: 10px; } .stat-row .stat-item { flex: 1; } </style> </head> <body> <div class="container"> <div class="header"> <h1> <span>❤️</span> 进程健康监控系统 </h1> <div class="status-bar"> <div class="status-indicator"> <div class="status-dot" id="statusDot"></div> <span id="connectionStatus">连接中...</span> </div> <span id="lastUpdate"></span> </div> </div> <div class="summary-cards"> <div class="card"> <h3>总进程数</h3> <div class="value" id="totalProcesses">0</div> </div> <div class="card healthy"> <h3>✓ 健康</h3> <div class="value" id="healthyCount">0</div> </div> <div class="card warning"> <h3>⚠ 警告</h3> <div class="value" id="warningCount">0</div> </div> <div class="card unhealthy"> <h3>✗ 异常</h3> <div class="value" id="unhealthyCount">0</div> </div> </div> <div class="processes-section"> <h2>进程详情</h2> <div id="processList"> <div class="no-data">等待数据...</div> </div> </div> <div class="last-update" id="footerUpdate"></div> </div> <script> let ws = null; let reconnectTimer = null; function connect() { ws = new WebSocket('ws://localhost:8282'); ws.onopen = function() { console.log('WebSocket 连接成功'); updateConnectionStatus(true); ws.send(JSON.stringify({ action: 'getHealth' })); }; ws.onmessage = function(event) { const message = JSON.parse(event.data); if (message.action === 'healthUpdate') { updateUI(message.data, message.timestamp); } }; ws.onclose = function() { console.log('WebSocket 连接关闭'); updateConnectionStatus(false); reconnectTimer = setTimeout(connect, 5000); }; ws.onerror = function(error) { console.error('WebSocket 错误:', error); }; } function updateConnectionStatus(connected) { const statusDot = document.getElementById('statusDot'); const statusText = document.getElementById('connectionStatus'); if (connected) { statusDot.classList.remove('disconnected'); statusText.textContent = '已连接'; } else { statusDot.classList.add('disconnected'); statusText.textContent = '连接断开'; } } function updateUI(data, timestamp) { // 更新汇总数据 document.getElementById('totalProcesses').textContent = data.summary.total; document.getElementById('healthyCount').textContent = data.summary.healthy; document.getElementById('warningCount').textContent = data.summary.warning; document.getElementById('unhealthyCount').textContent = data.summary.unhealthy; document.getElementById('lastUpdate').textContent = `最后更新: ${timestamp}`; document.getElementById('footerUpdate').textContent = `最后更新时间: ${timestamp}`; // 更新进程列表 const processList = document.getElementById('processList'); if (data.processes && data.processes.length > 0) { let html = ''; data.processes.forEach(process => { const statusIcon = getStatusIcon(process.status); const uptime = formatUptime(Date.now() / 1000 - process.start_time); const errorRate = process.requests > 0 ? ((process.errors / process.requests) * 100).toFixed(2) : 0; html += ` <div class="process-item ${process.status}"> <div class="process-header"> <div class="process-title"> <span class="status-icon">${statusIcon}</span> <span>PID ${process.pid} - Worker ${process.worker_id}</span> </div> <span class="status-badge ${process.status}">${getStatusText(process.status)}</span> </div> <div class="process-stats"> <div class="stat-item"> <div class="stat-label">CPU 使用率</div> ${renderProgressBar(process.cpu, '%')} </div> <div class="stat-item"> <div class="stat-label">内存使用率</div> ${renderProgressBar(process.memory, '%')} </div> <div class="stat-item"> <div class="stat-label">内存占用</div> <div class="stat-value">${process.memory_mb} MB</div> </div> <div class="stat-item"> <div class="stat-label">运行时长</div> <div class="stat-value">${uptime}</div> </div> <div class="stat-item"> <div class="stat-label">连接数</div> <div class="stat-value">${process.connections}</div> </div> <div class="stat-item"> <div class="stat-label">请求数</div> <div class="stat-value">${process.requests}</div> </div> <div class="stat-item"> <div class="stat-label">错误数</div> <div class="stat-value" style="color: ${process.errors > 0 ? '#ef4444' : '#22c55e'}">${process.errors}</div> </div> <div class="stat-item"> <div class="stat-label">错误率</div> <div class="stat-value" style="color: ${errorRate > 5 ? '#ef4444' : '#22c55e'}">${errorRate}%</div> </div> </div> ${process.warnings && process.warnings.length > 0 ? ` <div class="warnings"> <h4>⚠ 告警信息</h4> <ul> ${process.warnings.map(w => `<li>${w}</li>`).join('')} </ul> </div> ` : ''} </div> `; }); processList.innerHTML = html; } else { processList.innerHTML = '<div class="no-data">暂无进程数据</div>'; } } function getStatusIcon(status) { const icons = { 'healthy': '✅', 'warning': '⚠️', 'unhealthy': '❌' }; return icons[status] || '❓'; } function getStatusText(status) { const texts = { 'healthy': '健康', 'warning': '警告', 'unhealthy': '异常' }; return texts[status] || '未知'; } function renderProgressBar(value, unit) { let className = ''; if (value > 80) className = 'danger'; else if (value > 50) className = 'warning'; return ` <div class="progress-bar"> <div class="progress-fill ${className}" style="width: ${Math.min(value, 100)}%"> ${value}${unit} </div> </div> `; } function formatUptime(seconds) { const hours = Math.floor(seconds / 3600); const minutes = Math.floor((seconds % 3600) / 60); const secs = Math.floor(seconds % 60); return `${pad(hours)}:${pad(minutes)}:${pad(secs)}`; } function pad(num) { return num.toString().padStart(2, '0'); } // 页面加载时连接 connect(); // 页面关闭时断开连接 window.addEventListener('beforeunload', function() { if (ws) { ws.close(); } if (reconnectTimer) { clearTimeout(reconnectTimer); } }); </script> </body> </html>先运行php health_monitor.php start
3个评论
学习下
谢谢 我会继续努力的
不错，可以写成webman插件
谢谢 知道了
感谢分享
好