吴洪东-2018年7月11日记录-博客https://segmentfault.com/u/wu...Linux系统监控交互命令系统监控命令toptop:d输入数字n回车，表示按n刷新秒;空格键和回车键，表示立即刷新；q表示退出top监控；#调用toptop；#输出top-22:22:13up11:35,2users,loadaverage:0.02,0.03,0.05Tasks:108total,3running,105sleeping,0stopped,0zombie%Cpu(s):0.2us,0.4sy，0.0ni，99.2id，0.0wa，0.0hi，0.2si，0.0stKiBMem：总计3863544，3366220f，已使用167452，329872buff/cacheKiB交换：总计2097148，免费2097148，已使用0。3422540availMemPIDUSERPRNIVIRTRESSHRS%CPU%MEMTIME+COMMAND9root2006.000S2cu_0sched0.2:905root20026162883165176R0.20.21:05.62vmtoolsd解释：输出信息top——后面是当前时间，后面是系统启动运行的时间，有2个登录用户；loadaverage后面的三个值分别代表1分钟、5分钟、15分钟的系统平均负载；KiBMem后面是物理内存总量，free，和used；KiBSwap后面是虚拟内存的总量、空闲、已用；关注负载情况：一般看最近15分钟的负载，单核CPU如果达到1就说明负载已经到了；对于双核CPU，如果达到2，就说明已经达到了负载。然后一个一般当系统CPU核心数*0.7超过这个值，就得找系统问题了；命令iostat#installsysstatyuminstall-ysysstat;#调出I/O监控[root@masterwuhd]#iostatLinux3.10.0-862.2.3.el7.x86_64(master.wuhongdong.com)July11,2018_x86_64_(1CPU)avg-cpu:%user%nice%system%iowait%steal%idle0.180.000.280.010.0099.53Device:tpskB_read/skB_wrtn/skB_readkB_wrtnsda0.3510.291.4543739561528dm-00.369.581.4040739759458dm-10.000.050.0022480dm-20.000.010.005120[root@masterwuhd]#iostat-xLinux3.10.0-862.2。3.el7.x86_64(master.wuhongdong.com)July11,2018_x86_64_(1CPU)avg-cpu:%user%nice%system%iowait%steal%idle0.180.000.290.010.0099.53Device:rrqm/swrqm/sr/sw/srkB/swkB/savgrq-szavgqu-szawaitr_awaitw_awaitsvctm%utilsda0.000.020.200.1510.131.4965.780.001.950.813.501.090.04dm-00.000.000.190.179.431.4460.110.002.100.833.571.060.04dm-10.000.000.000.000.050.0045.410.000.160.160.000.150.00dm-20.000.000.000.000.010.0030.120.000.030.030.000.000.150.00dm-20.000.000.000.000.010.0030.120.000.030.030.000.000.00Inaperiodoftime,itmeans硬盘有严重瓶颈/值大于4iowait；在一段时间内，当%util的值接近100%时，表示设备带宽已满；
