tomcat6.0的大量CLOSE_WAIT解决方案

遇到tomcat closedwait相关解决方案

方案一，在tomcat的workers.properties里面配置相关条件

worker.tomcat.lbfactor=50
worker.tomcat.cachesize=10
worker.tomcat.cache_timeout=600
worker.tomcat.socket_keepalive=1
worker.tomcat.socket_timeout=300

方案二，优化服务器内核参数（/etc/sysctl.conf）
net.ipv4.tcp_tw_reuse=1
net.ipv4.tcp_tw_recycle=1
net.ipv4.tcp_keepalive_time=1800
net.ipv4.tcp_keepalive_probes=4
net.ipv4.tcp_keepalive_intvl=15

方案三，关闭tomcat的keepalive参数

方案四，做定时任务检查httpd进程数目，一旦发现httpd进程数目异常，就重启apache和tomcat

<strong> 针对于本项目目前的状况，我采用了方案四，编写了linux shell脚本如下：</strong>

#!/bin/sh
httpcount=`ps -ef|grep httpd|wc -l`
echo $httpcount
if [ $httpcount -gt 900 ]; then
  killall httpd
  killall java
  /cicro/wcm/bin/startas.sh
  echo "Restart successfully because of much useless process."
else
  echo "It is OK!"
fi

本文只是记录网上相关解决方案，原链 http://www.myexception.cn/operating-system/1276416.html