Nginx 缓存，redis_pass

Question

我整天都在用头撞墙。

我正在使用下面的 nginx 配置来测试一些东西

location  /help {
  set $redis_key "cache:$scheme://$host$request_uri";
  default_type   text/html;
  redis_pass 127.0.0.1:6379;
  error_page 404 = @upstream;

}

在我的 redis 实例中有一个用于缓存的键和值：$scheme....（在我的情况下是缓存：http://localhost/help）

我知道它们存在，因为我可以监视 redis-cli 的 nginx redis 请求，复制 "get" "cache:http://localhost/help"，将其粘贴到另一个 redis-cli 窗口并获得预期的响应。

问题来自 nginx，它没有得到响应。我再次可以看到它从redis-cli -> monitor 内部连接，并且我知道键和值存在。

从 nginx 错误日志我可以看到这个

2016/04/08 16:52:42 [notice] 9304#0: worker process 6328 exited with  code 0
2016/04/08 16:52:42 [notice] 9304#0: signal 29 (SIGIO) received
terminate called after throwing an instance of 'std::length_error'
  what():  basic_string::append
2016/04/08 16:52:49 [notice] 9304#0: signal 17 (SIGCHLD) received
2016/04/08 16:52:49 [alert] 9304#0: worker process 7328 exited on signal 6 (core dumped)
2016/04/08 16:52:49 [notice] 9304#0: start worker process 7516
2016/04/08 16:52:49 [notice] 9304#0: signal 29 (SIGIO) received
terminate called after throwing an instance of 'std::length_error'
  what():  basic_string::append
2016/04/08 16:52:50 [notice] 9304#0: signal 17 (SIGCHLD) received
2016/04/08 16:52:50 [alert] 9304#0: worker process 7335 exited on signal 6 (core dumped)
2016/04/08 16:52:50 [notice] 9304#0: start worker process 7544
2016/04/08 16:52:50 [notice] 9304#0: signal 29 (SIGIO) received
terminate called after throwing an instance of 'std::length_error'
  what():  basic_string::append

这是否与其他人有关，或者有人可以将我踢向正确的方向吗？

提前致谢

Answer 1

供以后阅读本文的任何人使用。

首先，过去的你好！

其次，原来nginx pagespeed模块和这种缓存是不兼容的。