Springboot 获取http数据、打印HTTP参数的4种方式 (便于生产排查问题)

Java的话本地打断点可以调试获取rest入参（http header），但是在生产环境可能我们获取入参（Http header/parameter）可能就没有那么的轻松了。我们可能在header中放置了很多自定的参数用来鉴权或者其他用途。如果排查问题的时候需要这些参数，我们有很多种选择去获取这些参数。

1. 输出到应用日志中，比如使用logback,log.error(xxx)
2. 借助nginx 输出到access.log日志中
3. 借助Skywalking/zipkin等中间件输出到链路中
4. 网关日志中输出

1. 输出到应用日志中

我们可以借助Springboot的拦截器在进入rest controller 之前将request header / param 输出出来，在rest controller调用结束之后将response header / param输出。

LogInterceptor
拦截器，注意拦截器和过滤器的区别，过滤器属于Tomcat/Jetty/… Servlet 容器的生命周期维护的，要早于拦截器。过滤器是Springboot维护的拦截，在handler mapping 映射之后先去调用拦截器之后在调用controller。

拦截器拦截的就是上图4的这部分。

@Component
@Slf4j
public class LogInterceptor extends HandlerInterceptorAdapter {

    private static ThreadLocal<Long> threadLocal = new ThreadLocal<>();

    @Override
    public boolean preHandle(HttpServletRequest request, HttpServletResponse response, Object handler) throws Exception {
        threadLocal.set(System.currentTimeMillis());
        log.info("Request uri = [{}], method is: [{}]", request.getRequestURI(), request.getMethod());
        log.info("Request header is : [{}]", parseRequestHeaders(request));
        log.info("Request param is : [{}]", parseParams(request));

        if (request instanceof RequestCustomWrapper) {
            RequestCustomWrapper requestCustomWrapper = (RequestCustomWrapper) request;
            byte[] body = requestCustomWrapper.getBody();
            log.info("Request body is : [{}]", new String(body));
        }

        return super.preHandle(request, response, handler);
    }

    public static String parseParams (HttpServletRequest request) {
        StringBuilder stringBuilder = new StringBuilder();
        Enumeration<String> parameterNames = request.getParameterNames();
        while (parameterNames.hasMoreElements()) {
            String name = parameterNames.nextElement();
            request.getParameter(name);
            stringBuilder.append(name).append("=").append(";");
        }
        return stringBuilder.toString();
    }

    public static String parseRequestHeaders (HttpServletRequest request) {
        StringBuilder stringBuilder = new StringBuilder();
        Enumeration<String> headerNames = request.getHeaderNames();
        while (headerNames.hasMoreElements()) {
            String name  = headerNames.nextElement();
            String value = request.getHeader(name);
            stringBuilder.append(name).append("=").append(value).append(";");
        }
        return stringBuilder.toString();
    }

    @Override
    public void afterCompletion(HttpServletRequest request, HttpServletResponse response, Object handler, Exception ex) throws Exception {
        super.afterCompletion(request, response, handler, ex);
    }
}

RequestCustomWrapper
因为Springboot框架规定,Request getInputStream只能获取一次，获取第二次的时候就会报错。所以这个时候需要实现RequestWrapper去包裹Request重写getInputStream实现可重复获取Request Stream。

@Slf4j
public class RequestCustomWrapper extends HttpServletRequestWrapper {

    private byte[] body;

    public byte[] getBody() {
        return body;
    }

    public RequestCustomWrapper(HttpServletRequest request) {
        super(request);
        try {
            body = readBytes(request.getReader());
        } catch (IOException e) {
            log.error("读取request input stream失败..");
        }
    }

    @Override
    public BufferedReader getReader() throws IOException {
        return new BufferedReader(new InputStreamReader(getInputStream()));
    }

    @Override
    public ServletInputStream getInputStream() throws IOException {
        try (final ByteArrayInputStream bais = new ByteArrayInputStream(body)) {
            return new ServletInputStream() {
                @Override
                public boolean isFinished() {
                    return false;
                }

                @Override
                public boolean isReady() {
                    return false;
                }

                @Override
                public void setReadListener(ReadListener readListener) {

                }

                @Override
                public int read() throws IOException {
                    return bais.read();
                }
            };
        }

    }

    public byte[] readBytes (BufferedReader br) throws IOException {
        byte[] emptyBytes = new byte[0];
        String str;
        StringBuilder sb = new StringBuilder();
        while ((str = br.readLine()) != null) {
            sb.append(str);
        }

        if (StringUtils.isNotBlank(sb.toString())) {
            return sb.toString().getBytes(StandardCharsets.UTF_8);
        }

        return emptyBytes;
    }
}

RequestCustomFilter
全局过滤器到，在执行到RequestCustomFilter这一层的时候，将ServletRequest包裹替换成自己的Request,实现可重复获取Request Stream.

public class RequestCustomFilter implements Filter {
    @Override
    public void doFilter(ServletRequest servletRequest, ServletResponse servletResponse, FilterChain filterChain) throws IOException, ServletException {
        if (servletRequest instanceof HttpServletRequest) {
            ServletRequest requestWrapper = new RequestCustomWrapper((HttpServletRequest) servletRequest);
            filterChain.doFilter(requestWrapper, servletResponse);
        } else {
            filterChain.doFilter(servletRequest, servletResponse);
        }
    }
}

WebMvcConfig
注册过滤器和拦截器

@Configuration
public class WebMvcConfig implements WebMvcConfigurer {

    @Override
    public void addInterceptors(InterceptorRegistry registry) {
        registry.addInterceptor(new LogInterceptor())
                .addPathPatterns("/**");
    }

    @Bean
    public FilterRegistrationBean<RequestCustomFilter> requestCustomFilter () {
        FilterRegistrationBean<RequestCustomFilter> registrationBean = new FilterRegistrationBean<>();
        RequestCustomFilter requestCustomFilter = new RequestCustomFilter();
        registrationBean.setFilter(requestCustomFilter);
        registrationBean.addUrlPatterns("/*");
        registrationBean.setOrder(1);
        return registrationBean;
    }
}

2.Nginx 配置输出Log

配置方式：
在nginx的配置文件中有个变量：$http_cookie来获取cookie的信息。配置方式很简单，只需要在nginx配置文件的http段，新添加一个log_format就可以了：
nginx.conf

log_format  sendfile  '$remote_addr - $remote_user [$time_local] "$request" '
              '$status $body_bytes_sent "$http_referer" '
              '"$http_user_agent" "$http_x_forwarded_for" "$http_cookie"';

在server.conf中加入

access_log  /var/log/php/access.log sendfile;

配置层级结构

sendfile 就是上面定义的log_format 的名字，只要acces_log 后面带这个名字的日志，就会按照定义的格式输出日志。

reload一下nginx就可以在日志里面看到cookie信息

nginx  -s  reload

Nginx 变量参考

1. $remote_addr #存放了客户端的地址，注意是客户端的公⽹IP
2. $args #变量中存放了URL中的指令http://www.magedu.net/main/index.do?id=090&partner=search以上：id=090&partner=search 即为 $args
3. $document_root #保存了针对当前资源的请求的系统根⽬录，如/apps/nginx/html
4. $cookie_name #表⽰key为 name 的cookie值
5. $document_uri #保存了当前请求中不包含指令的URI，注意是不包含请求的指令，如http://www.magedu.net/main/index.do?id=090&partner=search会被定义为/main/index.do
6. $host；#存放了请求的host名称
7. $http_user_agent #客户端浏览器的详细信息
8. $http_cookie #客户端的cookie信息
9. $limit_rate #如果nginx服务器使⽤limit_rate配置了显⽰⽹络速率，则会显⽰，如果没有设置， 则显⽰0
10. $remote_port #客户端请求Nginx服务器时客户端随机打开的端⼝
11. $remote_user #已经经过Auth Basic Module验证的⽤户名
12. $request_body_file #做反向代理时发给后端服务器的本地资源的名称
13. $request_method #请求资源的⽅式，GET/PUT/DELETE等
14. $request_filename #当前请求的资源⽂件的路径名称，由root或alias指令与URI请求⽣成的⽂件绝对路径，

3. 借助Skywalking/zipkin等中间件输出到链路中

请关注我，后续会单独出一期Skywalking的教程以及Skywalking原理讲解。

4. 网关日志中输出

这里只简单贴一下代码介绍网管如何打印Http信息，后续会单独出一期介绍网关的文章。请关注我！！！

@Component
@Slf4j
public class LoggingFilter implements GlobalFilter, Ordered {

    private static final String START_TIME = "START_TIME";

    @Override
    public Mono<Void> filter(ServerWebExchange exchange, GatewayFilterChain chain) {
        ServerHttpRequest request = exchange.getRequest();
        String info = String.format("Method: {%s} Host: {%s} Path: {%s} Query: {%s}",
                request.getMethod().name(),
                request.getURI().getHost(),
                request.getURI().getPath(),
                request.getQueryParams());
        log.info(info);
        exchange.getAttributes().put(START_TIME, System.currentTimeMillis());
        return chain.filter(exchange).then(Mono.fromRunnable(() -> {
            Long start = exchange.getAttribute(START_TIME);
            if (start != null) {
                long executeTime = System.currentTimeMillis() - start;
                log.info(exchange.getRequest().getURI().getRawPath() + ":" + executeTime + "ms");
            }
        }));
    }

    @Override
    public int getOrder() {
        return Ordered.LOWEST_PRECEDENCE;
    }
}