cgroup设置导致系统cpu过高?怎么解?

mopishv0 2013-05-14 11:41:35
之前希望用cgoup对hadoop进程的内存和使用那几块cpu进行限制
但是使用配置后
cpu load奇高 并且系统cpu占用都很高 正常进程几乎卡死
这是为什么呢?

mount {

cpuset = /cgroup/cpuset;

cpu = /cgroup/cpu;

cpuacct = /cgroup/cpuacct;

memory = /cgroup/memory;

devices = /cgroup/devices;

freezer = /cgroup/freezer;

net_cls = /cgroup/net_cls;

blkio = /cgroup/blkio;

}



group important {

cpuset {

cpuset.cpus = <%= cpus %>;

cpuset.mems = <%= mems %>;

} cpu {

cpu.shares = 768;

} memory {

memory.limit_in_bytes = 3072M;

}

}



group normal {

cpuset {

cpuset.cpus = <%= cpus %>;

cpuset.mems = <%= mems %>;

} cpu {

cpu.shares = 512;

} memory {

memory.limit_in_bytes = 800M;

}

}



group normal_1G {

cpuset {

cpuset.cpus = <%= cpus %>;

cpuset.mems = <%= mems %>;

} cpu {

cpu.shares = 512;

} memory {

memory.limit_in_bytes = 1024M;

}

}



group normal_2G {

cpuset {

cpuset.cpus = <%= cpus %>;

cpuset.mems = <%= mems %>;

} cpu {

cpu.shares = 512;

} memory {

memory.limit_in_bytes = 2048M;

}

}



group normal_3G {

cpuset {

cpuset.cpus = <%= cpus %>;

cpuset.mems = <%= mems %>;

} cpu {

cpu.shares = 512;

} memory {

memory.limit_in_bytes = 3072M;

}

}



group normal_4G {

cpuset {

cpuset.cpus = <%= cpus %>;

cpuset.mems = <%= mems %>;

} cpu {

cpu.shares = 512;

} memory {

memory.limit_in_bytes = 4096M;

}

}



group hadoop {

cpuset {

cpuset.cpus = <%= hadoop_service_cpus %>;

cpuset.mems = <%= mems %>;

}

}






以上是cgconfig的配置 <%= mems %>; <%= cpus %>; <%= hadoop_service_cpus %>;
会在puppet中被替换为具体指
求各位大大指教
很着急 可加分~谢谢
...全文
89 点赞 收藏 1
写回复
1 条回复
切换为时间正序
当前发帖距今超过3年,不再开放新的回复
发表回复
mopishv1 2013-05-15
http://us.generation-nt.com/answer/cgroup-scheduling-adding-kthreadd-non-rt-cgroup-can-deadlock-kernel-help-201640552.html 应该和这个有关 如果不设置 rt_runtime 值为0 导致其中说明的内核死锁 另 没有回复收分
回复
相关推荐
发帖
系统维护与使用区
创建于2007-08-27

1.9w+

社区成员

系统使用、管理、维护问题。可以是Ubuntu, Fedora, Unix等等
申请成为版主
帖子事件
创建了帖子
2013-05-14 11:41
社区公告
暂无公告