kernel: swapper: page allocation failure. order:1, mode:0x20

场景：领导电话通知，我们的主站宕机了，到家后从另外一台机器上ssh一直处于等待状态，开始怀疑机器的负载比较高，
后查看监控机器，发现网卡、cpu、nginx连接数.....通通都没有数据了，显然不是负载高度问题了，应该是机器死机了，立刻通过ipmi重启机器
重启机器后，机器正常！
其实这个机器都正常运行大半年了，没啥问题！ 
查询/var/log/messages，发现大量的信息如下：

点击(此处)折叠或打开

Mar 12 11:15:04 hy1 kernel: php-fpm: page allocation failure. order:1, mode:0x20
Mar 12 11:15:04 hy1 kernel: php-fpm: page allocation failure. order:1, mode:0x20
Mar 12 11:15:04 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:05 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:05 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:05 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:05 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:05 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:05 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:06 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:09 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:09 hy1 kernel: nginx: page allocation failure. order:1, mode:0x20
Mar 12 11:15:09 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:10 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:11 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:11 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:11 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:11 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:15:11 hy1 kernel: mysqld: page allocation failure. order:1, mode:0x20
Mar 12 11:17:33 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:53 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:53 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:53 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:53 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:54 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:54 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20
Mar 12 11:17:54 hy1 kernel: swapper: page allocation failure. order:1, mode:0x20

开始怀疑是系统的内存被吃光了，但通过检查监控，发现出问题的时候，内存还有蛮多可以用的！ 当时的内存使用情况，见附件！

后来查到是内核的的一个bug
解决方法如下：
vi /etc/sysctl.conf
写入：
vm.zone_reclaim_mode = 1
sysctl  -p  使其理解生效

 

原文：http://blog.chinaunix.net/uid-20776139-id-4155388.html