实验环境：centos 6.10
1.安装libcgroup
yum install -y libcgroup
2.进入资源控制器默认挂载目录/cgroup
[root@hadoop1 cgroup]# cd /cgroup/ [root@hadoop1 cgroup]# ll total 8 drwxr-xr-x. 2 root root 0 Nov 12 15:22 blkio drwxr-xr-x. 3 root root 0 Nov 12 16:22 cpu drwxr-xr-x. 2 root root 0 Nov 12 15:22 cpuacct drwxr-xr-x. 2 root root 0 Nov 12 15:22 cpuset drwxr-xr-x. 2 root root 0 Nov 12 15:22 devices drwxr-xr-x. 2 root root 0 Nov 12 15:22 freezer drwxr-xr-x. 3 root root 0 Nov 12 15:22 memory drwxr-xr-x. 2 root root 0 Nov 12 15:22 net_cls
3.创建两个测试脚本，use_cpu.sh模拟使用cpu的进程，use_mem.sh模拟使用内存的进程
[root@hadoop1 cgroup]# ll total 8 drwxr-xr-x. 2 root root 0 Nov 12 15:22 blkio drwxr-xr-x. 3 root root 0 Nov 12 16:22 cpu drwxr-xr-x. 2 root root 0 Nov 12 15:22 cpuacct drwxr-xr-x. 2 root root 0 Nov 12 15:22 cpuset drwxr-xr-x. 2 root root 0 Nov 12 15:22 devices drwxr-xr-x. 2 root root 0 Nov 12 15:22 freezer drwxr-xr-x. 3 root root 0 Nov 12 15:22 memory drwxr-xr-x. 2 root root 0 Nov 12 15:22 net_cls -rwxr-xr-x. 1 root root 49 Nov 12 15:28 use_cpu.sh -rwxr-xr-x. 1 root root 53 Nov 12 16:41 use_mem.sh [root@hadoop1 cgroup]# cat use_cpu.sh #!/bin/bash x=0 while [ true ]; do x=$x+1 done; [root@hadoop1 cgroup]# cat use_mem.sh #!/bin/bash x="a" while [ true ];do x=$x$x done;
4.直接运行脚本，分别查看cpu和内存的使用情况
[root@hadoop1 cgroup]# nohup ./use_cpu.sh & [3] 7817 [root@hadoop1 cgroup]# nohup: ignoring input and appending output to `nohup.out'
[root@hadoop1 cgroup]# top
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7817 root 20 0 103m 1724 1148 R 100.0 0.1 0:52.89 use_cpu.sh
[root@hadoop1 cgroup]# kill -9 7817
[root@hadoop1 cgroup]# nohup ./use_mem.sh &
[3] 7822
[root@hadoop1 cgroup]# nohup: ignoring input and appending output to `nohup.out'
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
7822 root 20 0 2151m 1.3g 1132 R 100.0 73.3 1:04.15 use_mem.sh
可以看到，cpu使用量几乎达到100%，内存使用量在不断增长，截图为止显示为73.3M。
5.创建一个cpu控制组，然后在其中运行任务，使用top观察结果
[root@hadoop1 cpu]# cgcreate -g cpu:foo [root@hadoop1 cpu]# echo 50000 > foo/cpu.cfs_quota_us # cpu.cfs_period_us中的值为100000，配置50000即使用50% [root@hadoop1 cgroup]# cgexec -g cpu:foo nohup ./use_cpu.sh & [3] 7832 [root@hadoop1 cgroup]# nohup: ignoring input and appending output to `nohup.out' PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 7832 root 20 0 103m 1500 1148 R 49.9 0.1 0:16.01 use_cpu.sh 462 root 20 0 0 0 0 S 0.3 0.0 0:01.96 jbd2/dm-0-8 1 root 20 0 19232 752 752 S 0.0 0.0 0:02.84 init 2 root 20 0 0 0 0 S 0.0 0.0 0:00.01 kthreadd 3 root RT 0 0 0 0 S 0.0 0.0 0:00.03 migration/0
可以看到，测试进程使用了约50%的cpu。同理测试控制内存，分配10M给进程组
[root@hadoop1 cgroup]# cgcreate -g memory:foo [root@hadoop1 cgroup]# echo 10485760 > memory/foo/memory.limit_in_bytes [root@hadoop1 cgroup]# cgexec -g memory:foo nohup ./use_mem.sh & [3] 7836 PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 7836 root 20 0 871m 11m 1156 R 87.1 0.6 1:07.82 use_mem.sh
计算可知，测试进程大约使用了10M内存
