The installation and use of substantially iostat

A, yum installation iostat
#yum install -y sysstat #注意不是安装iostat
Second, the method described (look at an example)

Options
-d: disk status display indication
-m: M represents a unit display in megabytes, -k display in bytes
1: represents a refresh frequency
3: represents 3 show, as in FIG
be directly keep in the rear of the device name such as # iostat -d sda -m, see a specified partition status
attribute :

TPS: the number of transfers per second per second transmission, the IO can be understood as one, but sometimes may also be a plurality of logical requests into the IO request
MB_read / s: second apparatus information read from
MB_wrtn / s: second information writing device
MB_read: the total amount of data read
MB_wrtn: write the total amount of data into
other options
-x: View more details
-c: view the status of the CPU case

property
rrqm / s: the read request related to equipment per second, how many were merge, the system calls need to read when when fetching data on the disk, VFS (virtual file system will read and write requests issued to each FS, FS was found when a plurality of requests to read a block when the same pieces, FS which will be merged)
wrqm / S: the second device-dependent, please write Seeking how much is incorporated
avgrq-sz: the average size of a sector request
avgqu-sz: average request queue length
await: the average time each IO request processing, the response time can be understood as the IO, IO response time of the system in general should be less than 5ms, 10ms is greater than if relatively large. This time includes the time and queue service time, that is to say, in general, larger than the await the svctm, the smaller the difference between them, then the queue shorter time, whereas the greater the difference, the longer the time the queue, the system described the problem
r_await: average time to process a read request of
w-await: the average write request processing time
svctm: average service time each IO operation of the device
% util: within the statistical time, all the IO processing time divided by the time statistics, it represents the equipment busy