iostat is an important tool used to monitor disk I/O in Linux. Installation requires the sysstat package; 1. Use iostat -d to view disk read and write status; 2. Use iostat -dx 2 5 to obtain extended statistics to judge performance bottlenecks; 3. Pay attention to key indicators such as %util and await; 4. Combining top/htop and long-term records for comprehensive analysis. Mastering its usage can help quickly locate disk-related performance issues.
Monitoring disk I/O is an important part of troubleshooting system performance issues, and iostat
is a very practical tool, especially suitable for use in Linux systems. It can help us quickly understand the disk's read and write status, load status, and whether there are bottlenecks.
Install the iostat tool
In most Linux distributions, iostat
is not installed by default, it belongs to the sysstat
package. You can install it through the system's package manager:
-
Debian/Ubuntu:
sudo apt install sysstat
CentOS/RHEL:
sudo yum install sysstat
After the installation is completed, you can directly run iostat
through the command line to check the disk status.
Basic usage and output interpretation
The simplest command is to enter directly:
iostat
This displays summary information on CPU usage and device I/O. But if you only care about the disk part, you can add the -d
parameter:
iostat -d
Several common key columns include:
- tps : Number of I/O requests transmitted per second (IO/s)
- kB_read/s and kB_wrtn/s : the amount of data read and write per second (unit KB)
- kB_read and kB_wrtn : The total amount of data read and written in total
This data can help you determine whether there are currently a large number of read or write operations in progress.
Specify intervals and times to observe real-time changes
To monitor disk I/O more dynamically, you can specify refresh intervals and times, such as refreshing every 2 seconds, showing 5 times in total:
iostat 2 5
This allows you to see the trends that change over time. If you find that the tps or service time (await) of a certain disk is significantly higher than that of other disks, there may be performance bottlenecks.
You can also add the -x
parameter to enable extended statistics:
iostat -dx 2 5
The more important fields are:
- %util : Device utilization, indicating the percentage of time the device processes I/O (if it exceeds 70%, you need to pay attention to it)
- await : Average waiting time (milliseconds) per I/O request. High value indicates slow response
- svctm : Actual service time (gradually deprecated)
These metrics combined can help you determine whether the disk has become a performance bottleneck.
Practical usage suggestions
In daily operation and maintenance, monitoring disk I/O is not just a "snap". Here are some practical suggestions:
- If you are running a database or log service, it is recommended to use
iostat -dx 1
to observe the real-time load regularly. - To view the CPU and process status with
top
orhtop
, it helps to comprehensively determine the source of performance bottlenecks. - For long-term monitoring, you can use scripts to record the output of
iostat
into a file for easier subsequent analysis. - Pay attention to distinguish the performance differences between mechanical hard drives and SSDs. SSDs are usually lower in await.
Basically that's it. Mastering the basic usage of iostat
can help you quickly locate many disk-related performance issues.
The above is the detailed content of How to monitor disk I O using iostat. For more information, please follow other related articles on the PHP Chinese website!

Hot AI Tools

Undress AI Tool
Undress images for free

Undresser.AI Undress
AI-powered app for creating realistic nude photos

AI Clothes Remover
Online AI tool for removing clothes from photos.

Clothoff.io
AI clothes remover

Video Face Swap
Swap faces in any video effortlessly with our completely free AI face swap tool!

Hot Article

Hot Tools

Notepad++7.3.1
Easy-to-use and free code editor

SublimeText3 Chinese version
Chinese version, very easy to use

Zend Studio 13.0.1
Powerful PHP integrated development environment

Dreamweaver CS6
Visual web development tools

SublimeText3 Mac version
God-level code editing software (SublimeText3)

When encountering Docker problems, you should first locate the problem, which is problems such as image construction, container operation or network configuration, and then follow the steps to check. 1. Check the container log (dockerlogs or docker-composelogs) to obtain error information; 2. Check the container status (dockerps) and resource usage (dockerstats) to determine whether there is an exception due to insufficient memory or port problems; 3. Enter the inside of the container (dockerexec) to verify the path, permissions and dependencies; 4. Review whether there are configuration errors in the Dockerfile and compose files, such as environment variable spelling or volume mount path problems, and recommend that cleanbuild avoid cache dryness

To manage Linux user groups, you need to master the operation of viewing, creating, deleting, modifying, and user attribute adjustment. To view user group information, you can use cat/etc/group or getentgroup, use groups [username] or id [username] to view the group to which the user belongs; use groupadd to create a group, and use groupdel to specify the GID; use groupdel to delete empty groups; use usermod-aG to add users to the group, and use usermod-g to modify the main group; use usermod-g to remove users from the group by editing /etc/group or using the vigr command; use groupmod-n (change name) or groupmod-g (change GID) to modify group properties, and remember to update the permissions of relevant files.

The steps to install Docker include updating the system and installing dependencies, adding GPG keys and repositories, installing the Docker engine, configuring user permissions, and testing the run. 1. First execute sudoaptupdate and sudoaptupgrade to update the system; 2. Install apt-transport-https, ca-certificates and other dependency packages; 3. Add the official GPG key and configure the warehouse source; 4. Run sudoaptinstall to install docker-ce, docker-ce-cli and containerd.io; 5. Add the user to the docker group to avoid using sudo; 6. Finally, dock

Adjusting kernel parameters (sysctl) can effectively optimize system performance, improve network throughput, and enhance security. 1. Network connection: Turn on net.ipv4.tcp_tw_reuse to reuse TIME-WAIT connection to avoid enabling tcp_tw_recycle in NAT environment; appropriately lower net.ipv4.tcp_fin_timeout to 15 to 30 seconds to speed up resource release; adjust net.core.somaxconn and net.ipv4.tcp_max_syn_backlog according to the load to cope with the problem of full connection queue. 2. Memory management: reduce vm.swappiness to about 10 to reduce

To restart the service managed by systemctl in Linux, 1. First use the systemctlstatus service name to check the status and confirm whether it is necessary to restart; 2. Use the sudosystemctlrestart service name command to restart the service, and ensure that there is administrator privileges; 3. If the restart fails, you can check whether the service name is correct, whether the configuration file is wrong, or whether the service is installed successfully; 4. Further troubleshooting can be solved by viewing the log journalctl-u service name, stopping and starting the service first, or trying to reload the configuration.

Bash scripts handle command line parameters through special variables. Use $1, $2, etc. to get positional parameters, where $0 represents the script name; iterates through "$@" or "$*", the former retains space separation, and the latter is merged into a single string; use getopts to parse options with parameters (such as -a, -b:value), where the option is added to indicate the parameter value; at the same time, pay attention to referring to variables, using shift to move the parameter list, and obtaining the total number of parameters through $#.

Managing server configuration is actually quite annoying, especially when there are more machines, it becomes unrealistic to manually modify configurations one by one. Chef is a tool that can help you handle these things automatically. With it, you can manage the state of different servers uniformly and make sure they all run the way you want. The key point is: write code to manage configuration, rather than typing commands by hand. 1. Don’t skip the installation and basic settings. The first step is to install the environment. You need to deploy ChefServer on a server, then install ChefClient on the managed node and complete the registration. This process is a bit like connecting a management center with its "little brother". The installation steps are roughly as follows: Install the ChefServer unit on the main control server

Software RAID can realize disk arrays through the operating system's own tools to improve performance or fault tolerance. 1. Use mdadm tools to create and manage RAID arrays under Linux, including installing, viewing hard disks, creating arrays, formatting, mounting and configuration saving; 2. Windows can realize the basic functions of RAID0 and RAID1 through "disk management", such as creating new strip volumes or mirrored volumes and formatting; 3. Notes include adding hot spare disks, monitoring the status regularly, high data recovery risks require backup, and the performance impacts that may be caused by certain levels.
