echo 15000000 > /proc/sys/kernel/sched_wakeup_granularity_ns
echo 10 > /proc/sys/kernel/sched_nr_migrate
echo 1000000 > /proc/sys/kernel/sched_migration_cost
linux调度器（九）——调度器的配置参数: http://blog.csdn.net/wudongxu/article/details/8574753
http://www.oschina.net/question/234345_48082
http://bbs.gfan.com/android-5357558-1-1.html
[root@dodge ~]# sysctl -A|grep "sched" | grep -v "domain"
kernel.sched_min_granularity_ns = 4000000
kernel.sched_latency_ns = 40000000
kernel.sched_wakeup_granularity_ns = 2000000
kernel.sched_batch_wakeup_granularity_ns = 25000000
kernel.sched_stat_granularity_ns = 0
kernel.sched_runtime_limit_ns = 40000000
kernel.sched_child_runs_first = 1
kernel.sched_features = 29
kernel.sched_compat_yield = 0
[root@dodge ~]#
http://artipc10.vub.ac.be/wordpress/2011/05/27/linux-performance-improvements/
CFS scheduler tuning
CFS (Competely Fair Scheduler) is the name of the Linux process scheduler. By default it is tuned for desktop workloads. For server systems where throughput is more important than latency, Red Hat’s tuned package proposes these sysctl settings for CFS for servers:
kernel.sched_min_granularity_ns = 10000000
kernel.sched_wakeup_granularity_ns = 15000000
https://www.kernel.org/doc/Documentation/scheduler/sched-design-CFS.txt
网站加速--服务器编写篇（上）: http://blog.sina.com.cn/s/blog_466c66400100bi2n.html
Linux技巧：多核下绑定硬件/进程到不同CPU: http://os.51cto.com/art/201007/212964.htm
http://wenku.baidu.com/link?url=OLCxGzfywyWlUp-1Zzs-