Three tools for performance monitoring in Linux

Three tools for performance monitoring in Linux
GuideHow to manage Linux Hosts? The 18 weapons can be regarded as Swiss Army knives. One or two weapons can be claimed. The most common one is the three brothers of the stat family.

What are the main computer resources? CPU, memory, and disk? Although there are many cloud computing technologies, it is still necessary to start investigating whether a system is running properly from these three aspects. Once the reason is found, you can change the program if you can change the program, if you cannot change the program, you will spend money to upgrade the hardware.

1. The boss mpstat is mainly used to view CPU reports, including user/system, iowait, and interruption.
2. The second is vmstat. Here m is not the Virtual Machine m, but the virtual memory m. Of course, it mainly depends on the memory (including cache and buffer.
3. The old three iostat is named as a hacker. In addition to generating disk usage information reports, it also includes CPU usage.

Here we mainly look at vmstat. Its Output content is roughly as follows (Forgive me, although it has not reached the limit, but I have tried a little bit to make this shell output look good, but failed ):

procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu----- r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st 0  0      0  96572 160052 119776    0    0    11    15    0    1  0  0 100  0  0

The name of the top-level category of the output is mainly the processor, memory, swap zone, io, system, and CPU usage. This is also the data content that can be seen by the vmstat command. The second line is the name of each project. For example, bi and bo in io indicate block in and block out per second, respectively, that is, the amount of data read from and written to block devices. For details about each item, refer to the man manual. If you do not know the specific content, you will naturally not need to find anyone you know.

Based on some information on the Internet, the following summarizes some simple "routines" for investigating system performance bottlenecks ". Of course, if you want to understand the following "routine", you have to understand the "black talk", that is, the meaning of the Project names output by the vmstat command.

From: http://mp.weixin.qq.com/s? _ Biz = MzI3OTEzNjI1OQ = &

Address: http://www.linuxprobe.com/linux-three-saber.html ghost