nginx的Connection refused

问题描述
nginx的错误日志中突然出现大量的的Connection refused问题，日志如下：

2020/03/19 09:52:53 [error] 20117#20117: *7403411764 connect() failed (111: Connection refused) while connecting to upstream, client: xxx.xxx.xxx.xxx, server: , request: "POST /post/result/lol?type=Bet HTTP/1.1", upstream: "http://xxx.xxx.xxx.xxx/post/result/lol?type=Bet", host: "xxx.xxx.xxx.xxx"
2020/03/19 09:52:53 [error] 20117#20117: *7403411774 connect() failed (111: Connection refused) while connecting to upstream, client: xxx.xxx.xxx.xxx, server: , request: "POST /post/result/csgo?type=RollingBet HTTP/1.1", upstream: "http://xxx.xxx.xxx.xxx/post/result/csgo?type=RollingBet", host: "xxx.xxx.xxx.xxx"
2020/03/19 09:52:54 [error] 20116#20116: *7403411815 connect() failed (111: Connection refused) while connecting to upstream, client: xxx.xxx.xxx.xxx, server: , request: "POST /post/result/lol?type=Bet HTTP/1.1", upstream: "http://xxx.xxx.xxx.xxx/post/result/lol?type=Bet", host: "xxx.xxx.xxx.xxx"

出现这个问题，一开始以为是server节点挂掉，但是查看了下server运行正常；这个错误是突然间爆发大量的错误，查看了相关nginx和服务器监控系统，看到连接数突增。可以说明在高负载下，系统响应变慢，并出现超时或失误失败情况，TIME_WAIT积压。

问题定位
查看了tcp连接命令

# netstat -n | awk '/^tcp/ {++S[$NF]} END {for(a in S) print a, S[a]}'
TIME_WAIT 35423
CLOSE_WAIT 23602
SYN_SENT 62
FIN_WAIT1 61
FIN_WAIT2 259
ESTABLISHED 7543
SYN_RECV 3
CLOSING 35
LAST_ACK 507

发现WAIT数量过高，TCP连接断开后，会以TIME_WAIT状态保留一定的时间，然后才会释放端口。当并发请求过多的时候，就会产生大量的TIME_WAIT状态的连接，无法及时断开的话，会占用大量的端口资源和服务器资源，导致很多连接被拒绝了。

修改系统参数

# vim /etc/sysctl.conf
net.ipv4.tcp_fin_timeout = 30 #保留 FIN_WAIT2 的时间, 默认值是60, 单位是秒.
net.ipv4.tcp_timestamps = 1 #时间戳可以避免序列号的卷绕，默认为0，表示关闭；
net.ipv4.tcp_tw_reuse = 1 #表示开启重用。允许将TIME-WAIT sockets重新用于新的TCP连接，默认为0，表示关闭；
net.ipv4.tcp_tw_recycle = 1 # 表示开启TCP连接中TIME-WAIT sockets的快速回收，默认为0，表示关闭。

配置生效

/sbin/sysctl -p

WAIT的数量降低了，nginx也没有报Connection refused。

# netstat -n | awk '/^tcp/ {++S[$NF]} END {for(a in S) print a, S[a]}'
TIME_WAIT 2521
CLOSE_WAIT 13602