> For the complete documentation index, see [llms.txt](https://oss-docs.gitbook.io/openresty-best-practices/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://oss-docs.gitbook.io/openresty-best-practices/ngx_lua/log.md).

# 正确的记录日志

看过本章第一节的同学应该还记得，`log_by_lua*` 是一个请求经历的最后阶段。由于记日志跟应答内容无关，Nginx 通常在结束请求之后才更新访问日志。由此可见，如果我们有日志输出的情况，最好统一到 `log_by_lua*` 阶段。如果我们把记日志的操作放在 `content_by_lua*` 阶段，那么将线性的增加请求处理时间。

在公司某个定制化项目中，Nginx 上的日志内容都要输送到 syslog 日志服务器。我们使用了 [lua-resty-logger-socket](https://github.com/cloudflare/lua-resty-logger-socket) 这个库。

> 调用示例代码如下（有问题的）：

```lua
-- lua_package_path "/path/to/lua-resty-logger-socket/lib/?.lua;;";
--
--    server {
--        location / {
--            content_by_lua_file lua/log.lua;
--        }
--    }

-- lua/log.lua
local logger = require("resty.logger.socket")
if not logger.initted() then
    local ok, err = logger.init{
        host = 'xxx',
        port = 1234,
        flush_limit = 1,   --日志长度大于 flush_limit 的时候会将 msg 信息推送一次
        drop_limit = 99999,
    }
    if not ok then
        ngx.log(ngx.ERR, "failed to initialize the logger: ", err)
        return
    end
end

local msg = string.format(.....)
local bytes, err = logger.log(msg)
if err then
    ngx.log(ngx.ERR, "failed to log message: ", err)
    return
end
```

在实测过程中我们发现了些问题：

* 缓存无效：如果 `flush_limit` 的值稍大一些（例如 2000），会导致某些体积比较小的日志出现莫名其妙的丢失，所以我们只能把 `flush_limit` 调的很小
* 自己拼写 msg 所有内容，比较辛苦

那么我们来看 [lua-resty-logger-socket](https://github.com/cloudflare/lua-resty-logger-socket) 这个库的 `log` 函数是如何实现的呢，代码如下：

```lua
function _M.log(msg)
    ...

    if (debug) then
        ngx.update_time()
        ngx_log(DEBUG, ngx.now(), ":log message length: " .. #msg)
    end

    local msg_len = #msg

    if (is_exiting()) then
        exiting = true
        _write_buffer(msg)
        _flush_buffer()
        if (debug) then
            ngx_log(DEBUG, "Nginx worker is exiting")
        end
        bytes = 0
    elseif (msg_len + buffer_size < flush_limit) then  -- 历史日志大小+本地日志大小小于推送上限
        _write_buffer(msg)
        bytes = msg_len
    elseif (msg_len + buffer_size <= drop_limit) then
        _write_buffer(msg)
        _flush_buffer()
        bytes = msg_len
    else
        _flush_buffer()
        if (debug) then
            ngx_log(DEBUG, "logger buffer is full, this log message will be " .. "dropped")
        end
        bytes = 0
        --- this log message doesn't fit in buffer, drop it

        ...
```

由于在 `content_by_lua*` 阶段变量的生命周期会随着请求的终结而终结，所以当日志量小于 `flush_limit` 的情况下这些日志就不能被累积，也不会触发 `_flush_buffer` 函数，所以小日志会丢失。

这些坑回头来看是这么明显，所有的问题都是因为我们把 `lua/log.lua` 用错阶段了，应该放到 `log_by_lua*` 阶段，所有的问题都不复存在。

> 修正后：

```
lua_package_path "/path/to/lua-resty-logger-socket/lib/?.lua;;";

server {
    location / {
        content_by_lua_file lua/content.lua;
        log_by_lua_file lua/log.lua;
    }
}
```

这里有个新问题，如果我的 log 里面需要输出一些 content 的临时变量，两阶段之间如何传递参数呢？

> 方法肯定有，推荐下面这个：

```
location /test {
    rewrite_by_lua_block {
        ngx.say("foo = ", ngx.ctx.foo)
        ngx.ctx.foo = 76
    }
    access_by_lua_block {
        ngx.ctx.foo = ngx.ctx.foo + 3
    }
    content_by_lua_block {
        ngx.say(ngx.ctx.foo)
    }
}
```

更多有关 ngx.ctx 信息，请看[这里](https://github.com/openresty/lua-nginx-module#ngxctx)。


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://oss-docs.gitbook.io/openresty-best-practices/ngx_lua/log.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.