wrk 压测中请求无法响应问题解决过程

================= 遇到问题 =================
$ 直连压测

wrk -c10000 -t100 -d100m http://localhost:9981/order/list

$ nginx代理压测

wrk -c10000 -t100 -d100m http://localhost:8864/order/list

压测外的请求无法响应

================= 搜素方案 =================
$ 修改linux参数配置(最大打开文件数与进程最多打开文件数)以支持大量连接
$ nginx压测时会产生大量time_wait,与后端通信时开启keep alive以缓解time_wait情况

================= 用到的工具 =================
// 查看linux tcp连接情况

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

// 更快查看linux tcp连接情况的方式

ss -ant | awk 'NR>1 {++s[$1]} END {for(k in s) print k,s[k]}';

// 每秒输出一次当前系统连接状态

ss.sh
#!/bin/bash
while :
do
ss -ant | awk 'NR>1 {++s[$1]} END {for(k in s) print k,s[k]}'
sleep 1s;
done

================= 最终解决 =================
$ 修改linux参数配置

wrk 压测中请求无法响应问题解决过程

modprobe ip_conntrack

vi /etc/sysctl.conf
fs.file-max = 1024000
net.netfilter.nf_conntrack_max = 1024000
net.nf_conntrack_max = 1024000
/sbin/sysctl -p vi /etc/security/limits.conf
* hard nofile 1024000
* soft nofile 1024000 try
net.netfilter.nf_conntrack_max = 1024000
net.nf_conntrack_max = 1024000
instead
net.ipv4.ip_conntrack_max = 1024000
net.ipv4.netfilter.ip_conntrack_max = 1024000

$ 修改nginx配置

/etc/nginx/nginx.conf
user nginx;
worker_processes 4; error_log /var/log/nginx/error.log warn;
pid /var/run/nginx.pid; events {
worker_connections 10240;
} http {
include /etc/nginx/mime.types;
default_type application/octet-stream; log_format main '$remote_addr - $remote_user [$time_local] "$request" '
'$status $body_bytes_sent "$http_referer" '
'"$http_user_agent" "$http_x_forwarded_for"'; access_log /var/log/nginx/access.log main; sendfile on;
#tcp_nopush on; keepalive_timeout 65; #gzip on; include /etc/nginx/conf.d/*.conf;
} /etc/nginx/conf.d/default.conf
upstream http_backend {
server 127.0.0.1:9981;
keepalive 512;
} server {
listen 8864;
server_name localhost; location /nginx {
default_type application/json;
return 200 '{"success": true}';
} location / {
proxy_pass http://http_backend;
proxy_http_version 1.1;
proxy_set_header Connection "";
} #error_page 404 /404.html; # redirect server error pages to the static page /50x.html
#
error_page 500 502 503 504 /50x.html;
location = /50x.html {
root /usr/share/nginx/html;
}
} nginx -s reload

$ 再次压测,直连和nginx代理都能正常响应(虽然很慢...应该与硬件条件有关)

参考文章:

构建C1000K的服务器(1) – 基础

构建C1000K的服务器(2) – 实现百万连接的comet服务器

sysctl: cannot stat /proc/sys/net/ipv4/netfilter/ip_conntrack_max: No such file or directory

上一篇:junit源码解析--核心类


下一篇:SSRF和XSS-filter_var(), preg_match() 和 parse_url()绕过学习