Laravel队列调用API遇cURL重试失败异常求助

问题场景 #

在Laravel项目中，我使用Http facade调用多个第三方API，并将这些API调用逻辑封装到队列任务（Job）中执行。本地开发环境测试完全正常，但部署到预发布服务器后，执行queue:work时触发以下cURL错误：

GuzzleHttp\Exception\RequestException: cURL error 0: The cURL request was retried 3 times and did not succeed. The most likely reason for the failure is that cURL was unable to rewind the body of the request and subsequent retries resulted in the same error. Turn on the debug option to see what went wrong. See https://bugs.php.net/bug.php?id=47204 for more information. (see https://curl.haxx.se/libcurl/c/libcurl-errors.html) for https://app.oncrawl.com/api/v2/projects?limit=100&offset=0&sort=name%3Adesc in /var/lib/jenkins/workspace/sonar-iq/vendor/guzzlehttp/guzzle/src/Handler/CurlFactory.php:276

当前Http facade的配置如下：

'verify' => false,
'timeout' => 0,
'debug' => true,
'curl' => [
    CURLOPT_FORBID_REUSE => false,
    CURLOPT_FRESH_CONNECT => true
]

关键现象：只有通过队列queue:work执行任务时才会报错，直接调用API逻辑完全正常。

可行解决方法 #

1. 禁用Guzzle自动重试机制 #

队列常驻进程环境下，Guzzle的默认重试逻辑会尝试重新发送请求，但如果请求体无法被重写（比如流已耗尽）就会触发这个错误。可以直接关闭重试：

$response = Http::withOptions([
    'verify' => false,
    'timeout' => 0,
    'curl' => [
        CURLOPT_FORBID_REUSE => false,
        CURLOPT_FRESH_CONNECT => true
    ],
    // 新增：关闭自动重试
    'retries' => 0
])->get('https://app.oncrawl.com/api/v2/projects?limit=100&offset=0&sort=name%3Adesc');

如果上述方式不生效，还可以通过HandlerStack移除重试中间件：

use GuzzleHttp\HandlerStack;

$stack = HandlerStack::create();
// 移除重试中间件
$stack->remove('retry');

$response = Http::withOptions([
    'verify' => false,
    'timeout' => 0,
    'curl' => [
        CURLOPT_FORBID_REUSE => false,
        CURLOPT_FRESH_CONNECT => true
    ],
    'handler' => $stack
])->get('https://app.oncrawl.com/api/v2/projects?limit=100&offset=0&sort=name%3Adesc');

2. 确保请求体可重绕 #

如果业务需要保留重试逻辑，要确保请求体是可重复读取的流。对于你的GET请求场景，可以手动构造可重绕的请求对象：

use GuzzleHttp\Psr7\Request;
use GuzzleHttp\Psr7\Utils;

$http = Http::withOptions([
    'verify' => false,
    'timeout' => 0,
    'curl' => [
        CURLOPT_FORBID_REUSE => false,
        CURLOPT_FRESH_CONNECT => true
    ]
]);

// 构造带可重绕空流的请求（GET请求通常无请求体，这里强制设置避免重写失败）
$request = new Request(
    'GET',
    'https://app.oncrawl.com/api/v2/projects?limit=100&offset=0&sort=name%3Adesc',
    [],
    Utils::streamFor('')
);

$response = $http->send($request);

3. 调整cURL连接复用配置 #

当前配置中CURLOPT_FRESH_CONNECT => true会强制每次新建连接，结合队列常驻进程可能导致资源释放异常。尝试调整这两个cURL选项：

'curl' => [
    CURLOPT_FORBID_REUSE => true,  // 禁止复用连接，避免残留状态
    CURLOPT_FRESH_CONNECT => false // 关闭强制新建连接，让cURL自行管理连接池
]

4. 队列任务中每次创建新的Http实例 #

队列Worker是常驻进程，全局Http facade可能存在状态残留问题。在任务的handle方法里每次都创建新的Http实例，避免复用：

use Illuminate\Http\Client\Factory;

public function handle()
{
    // 每次执行任务都新建Http实例
    $http = new Factory();
    $response = $http->withOptions([
        'verify' => false,
        'timeout' => 0,
        'curl' => [
            CURLOPT_FORBID_REUSE => false,
            CURLOPT_FRESH_CONNECT => true
        ]
    ])->get('https://app.oncrawl.com/api/v2/projects?limit=100&offset=0&sort=name%3Adesc');
}

内容的提问来源于stack exchange，提问作者Shabeersha Salim