今天服务器又宕了,ssh无法链接,服务也无法访问。重启后发现/var/log/message中错误日志:
Oct 20 13:00:55 uhz000143 automount[2681]: expire_proc: expire thread create for /net failed
Oct 20 13:01:48 uhz000143 automount[2681]: expire_proc: expire thread create for /misc failed
Oct 20 13:02:10 uhz000143 automount[2681]: expire_proc: expire thread create for /net failed
Oct 20 13:03:03 uhz000143 automount[2681]: expire_proc: expire thread create for /misc failed
Oct 20 13:03:25 uhz000143 automount[2681]: expire_proc: expire thread create for /net failed
Oct 20 13:04:18 uhz000143 automount[2681]: expire_proc: expire thread create for /misc failed
Oct 20 13:04:40 uhz000143 automount[2681]: expire_proc: expire thread create for /net failed
Oct 20 13:05:33 uhz000143 automount[2681]: expire_proc: expire thread create for /misc failed
Oct 20 13:05:55 uhz000143 automount[2681]: expire_proc: expire thread create for /net failed
网上说是系统线程总数超过系统限制所致,cat /proc/sys/kernel/threads-max 后看到系统内核的值是135168,虽然用了很多线程池,但貌似也没用到这么多线程啊,这个事情在7月份也发生过,中间莫名就好了,现在又出现了,是地雷总会被踩到...大家支个招帮帮我查查问题的原因所在?一般你们碰到这样的问题是怎样一个排查的思路。网上一篇关于“ExecutorService”GC的文章我看过了,除此之外,大家有什么别的建议?我的环境是centos5.4,java1.6,tomcat6,mysql5.5.8,RabbitMQ 2.8.4,apache2.2.memcached1.4.5。。。