ngx_lua_API 详解 (一)

ngx.thread.spawn、ngx.thread.wait、ngx.thread.kill 介绍

ngx_lua中访问多个第三方服务

ngx_lua中提供了ngx.socket API，可以方便的访问第三方网络服务。如下面的代码,通过get_response函数从两个（或者更多）的源服务器获取数据，再生成响应发给客户端。

location / {
    content_by_lua_block {
        local get_response(host, port)
            local sock = ngx.socket.tcp()
            local ok, err = sock:connect(host, port)
            if not ok then
                return nil, err
            end
            local data, err = sock:receive()
            if not data then
                return nil, err
            end

            return data
        end

        local first = get_response("lua.org", 8080)
        local second = get_response("nginx.org", 8080)
        ngx.say(first .. second)
    }
}      

如果需要10个第三方网络服务，需要调用get_response 10次。总的响应时间与需要连接源的数量成正比。那么如何缩短源的响应时间呢？ngx.thread就是用来解决这种问题的。

二、lua-nginx-module提供了三个API

　　1、ngx.thread.spawn

　　2、ngx.thread.wait

　　3、ngx.thread.kill

三、详解

1、ngx.thread.spawn

重点：ngx.thread.spawn生成新的"light thread"，这个"light thread"运行优先级比它的父协程高，会优先运行，父协程被迫暂停。"light thread"运行结束或者yield后，再由ngx_http_lua_run_posted_threads去运行父协程。

​​参考：ngx_lua中的协程调度(六)之ngx_http_lua_run_posted_thread​​

通过ngx.thread.spawn可以生成一个"light thread"，一个”light thread“和Lua的协程类似，区别在于"light thread"是由ngx_lua模块进行调度的，多个"light thread"同时运行。

"light thread"，协程 和 进程。"light thread"比Lua中的协程更像操作系统中的进程。

• fork生成新的进程，生成的多个进程可以同时运行，而ngx.thread.spawn生成新的协程，多个协程同时在跑。
• kill可以杀死不需要的子进程，ngx.thread.kill可以杀死不需要的"light thread"
• wait可以等待子进程结束并取得子进程退出状态，ngx.thread.wait可以等待"light thread"结束并获取其返回值。

ngx.thread的使用，用ngx.thread重写上面的代码

location / {
    content_by_lua_block {
        local get_response(host, port)
            local sock = ngx.socket.tcp()
            local ok, err = sock:connect(host, port)
            if not ok then
                return nil, err
            end
            local data, err = sock:receive()
            if not data then
                return nil, err
            end

            return data
        end

        local t1 = ngx.thread.spawn(get_response, "lua.org", 8080)
        local t2 = ngx.thread.spawn(get_response, "nginx.org", 8080)
        local ok, res1, res2 = ngx.thread.wait(t1, t2)
        ngx.say(res1 .. res2)
    }
}      

生成的两个"light thread"可以同时运行，总的耗时只相当于访问一个源服务器的时间，即使需要访问的源服务器增加，耗时没有太大的变化。

"light thread"的调度

　　Linux中的fork生成新的子进程，父进程与子进程谁先运行呢？都有可能，和系统的调度有关。

　　把调用ngx.thread.spawn的这个Lua协程称为父协程，生成的"light thread"和父协程谁先运行呢? 在ngx_lua的调度逻辑中，是生成的"light thread"先运行，运行结束或者被挂起后，父协程才会继续运行。实际的代码在ngx_http_lua_run_thread函数中，这个函数比较复杂，涉及的东西太多，稍后再细说。

如下面的代码，没有调用ngx.thread.wait去等待"light thread"的结束。

# [1] 没有调用ngx.thread.wait去等待"light thread"的结束
location /thread002 {
    content_by_lua_block {
        local function f(name)
            ngx.log(ngx.ERR, "thread name: ", name, ", now start")
            ngx.sleep(4)
            ngx.log(ngx.ERR, "thread name: ", name, ", now end")
        end
        local t1 = ngx.thread.spawn(f, "first")
        local t2 = ngx.thread.spawn(f, "second")
        ngx.log(ngx.ERR, "main thread end")
    }
}      

由Nginx的日志中可以看到当前的请求一直延迟到t1,t2两个"light thread"最后退出才会结束。 Nginx中日志的顺序也可以看出父协程和两个"light thread"的执行那个顺序。

2017/07/21 09:45:07 [error] 115387#0: *26 [lua] content_by_lua(thread.conf:140):3: thread name: first, now start, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread002 HTTP/1.1", host: "127.0.0.1:8689"
2017/07/21 09:45:07 [error] 115387#0: *26 [lua] content_by_lua(thread.conf:140):3: thread name: second, now start, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread002 HTTP/1.1", host: "127.0.0.1:8689"
2017/07/21 09:45:07 [error] 115387#0: *26 [lua] content_by_lua(thread.conf:140):10: main thread end, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread002 HTTP/1.1", host: "127.0.0.1:8689"
2017/07/21 09:45:11 [error] 115387#0: *26 [lua] content_by_lua(thread.conf:140):5: thread name: first, now end, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread002 HTTP/1.1", host: "127.0.0.1:8689"
2017/07/21 09:45:11 [error] 115387#0: *26 [lua] content_by_lua(thread.conf:140):5: thread name: second, now end, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread002 HTTP/1.1", host: "127.0.0.1:8689"      

而如果代码中主动调用了ngx.exit()结束请求，那么t1，t2两个没有打印出完全的信息就被kill掉了。

# [2] 代码中主动调用了ngx.exit()结束请求，那么t1，t2两个没有打印出完全的信息就被kill掉了
location /thread003 {
    content_by_lua_block {
        local function f(name)
            ngx.log(ngx.ERR, "thread name: ", name, ", now start")
            ngx.sleep(4)
            ngx.log(ngx.ERR, "thread name: ", name, ", now end")
        end
        local t1 = ngx.thread.spawn(f, "first")
        local t2 = ngx.thread.spawn(f, "second")
        ngx.exit(200)
    }
}      

相应的Nginx日志

2017/07/21 09:48:08 [error] 115508#0: *28 [lua] content_by_lua(thread.conf:156):3: thread name: first, now start, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread003 HTTP/1.1", host: "127.0.0.1:8689"
2017/07/21 09:48:08 [error] 115508#0: *28 [lua] content_by_lua(thread.conf:156):3: thread name: second, now start, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread003 HTTP/1.1", host: "127.0.0.1:8689"
2017/07/21 09:48:08 [error] 115508#0: *28 [lua] content_by_lua(thread.conf:156):10: main thread end, client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread003 HTTP/1.1", host: "127.0.0.1:8689"      

"light thread"的限制

2017/07/21 09:14:10 [error] 114598#0: *1 failed to load inlined Lua code: content_by_lua(thread.conf:128):2: ')' expected near ',', client: 127.0.0.1, server: 127.0.0.1, request: "GET /thread001 HTTP/1.1", host: "127.0.0.1:8689"