iostat command
Display CPU and I/O statistics for devices and partitions.
Overview
iostat reports CPU statistics and input/output statistics for devices, partitions, and network filesystems. It's primarily used for monitoring system input/output device loading by observing the time devices are active in relation to their average transfer rates.
Options
-c
Display only CPU statistics.
$ iostat -c
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
-d
Display only device statistics.
$ iostat -d
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 5.73 141.61 45.28 0.00 1250432 399764 0
sdb 0.02 0.63 0.00 0.00 5548 0 0
-x
Display extended statistics for devices.
$ iostat -x
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device r/s w/s rkB/s wkB/s rrqm/s wrqm/s %rrqm %wrqm r_await w_await aqu-sz rareq-sz wareq-sz svctm %util
sda 4.21 1.52 141.61 45.28 0.10 0.57 2.33 27.29 0.63 2.38 0.01 33.63 29.72 0.28 0.16
sdb 0.02 0.00 0.63 0.00 0.00 0.00 0.00 0.00 0.71 0.00 0.00 31.53 0.00 0.57 0.00
-k
Display statistics in kilobytes per second.
$ iostat -k
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 5.73 141.61 45.28 0.00 1250432 399764 0
sdb 0.02 0.63 0.00 0.00 5548 0 0
-m
Display statistics in megabytes per second.
$ iostat -m
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device tps MB_read/s MB_wrtn/s MB_dscd/s MB_read MB_wrtn MB_dscd
sda 5.73 0.14 0.04 0.00 1221 390 0
sdb 0.02 0.00 0.00 0.00 5 0 0
-p [device]
Display statistics for block devices and all their partitions.
$ iostat -p sda
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 5.73 141.61 45.28 0.00 1250432 399764 0
sda1 0.01 0.32 0.00 0.00 2832 0 0
sda2 5.71 141.29 45.28 0.00 1247600 399764 0
-t
Print the time for each report displayed.
$ iostat -t
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
05/05/2025 10:15:30 AM
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 5.73 141.61 45.28 0.00 1250432 399764 0
sdb 0.02 0.63 0.00 0.00 5548 0 0
interval [count]
Specify the reporting interval in seconds and the number of reports.
$ iostat 2 3
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 5.73 141.61 45.28 0.00 1250432 399764 0
sdb 0.02 0.63 0.00 0.00 5548 0 0
avg-cpu: %user %nice %system %iowait %steal %idle
3.50 0.00 1.75 0.25 0.00 94.50
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 10.50 320.00 112.00 0.00 640 224 0
sdb 0.00 0.00 0.00 0.00 0 0 0
avg-cpu: %user %nice %system %iowait %steal %idle
3.38 0.00 1.50 0.12 0.00 95.00
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 8.00 256.00 64.00 0.00 512 128 0
sdb 0.00 0.00 0.00 0.00 0 0 0
Usage Examples
Monitoring disk I/O with extended statistics
$ iostat -xd 5
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
Device r/s w/s rkB/s wkB/s rrqm/s wrqm/s %rrqm %wrqm r_await w_await aqu-sz rareq-sz wareq-sz svctm %util
sda 4.21 1.52 141.61 45.28 0.10 0.57 2.33 27.29 0.63 2.38 0.01 33.63 29.72 0.28 0.16
sdb 0.02 0.00 0.63 0.00 0.00 0.00 0.00 0.00 0.71 0.00 0.00 31.53 0.00 0.57 0.00
[Output repeats every 5 seconds]
Monitoring specific partitions
$ iostat -p sda 2
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
Device tps kB_read/s kB_wrtn/s kB_dscd/s kB_read kB_wrtn kB_dscd
sda 5.73 141.61 45.28 0.00 1250432 399764 0
sda1 0.01 0.32 0.00 0.00 2832 0 0
sda2 5.71 141.29 45.28 0.00 1247600 399764 0
[Output repeats every 2 seconds]
Displaying CPU and disk statistics in megabytes
$ iostat -cm 1 3
Linux 5.15.0-91-generic (hostname) 05/05/2025 _x86_64_ (8 CPU)
avg-cpu: %user %nice %system %iowait %steal %idle
3.25 0.00 1.42 0.35 0.00 94.98
[Output repeats 3 times at 1-second intervals]
Tips
Understanding %util
The %util metric shows the percentage of CPU time during which I/O requests were issued to the device. A value close to 100% indicates saturation, meaning the device is working at its maximum capacity.
Identifying I/O Bottlenecks
Look for high values in the await column (average time for I/O requests to be served). High await times combined with high %util values indicate potential I/O bottlenecks.
Continuous Monitoring
For real-time monitoring, use iostat with an interval (e.g., iostat -x 2). This will continuously update the statistics every 2 seconds until interrupted with Ctrl+C.
Combining with Other Tools
Use iostat in conjunction with tools like top, vmstat, and sar for a comprehensive system performance analysis.
Frequently Asked Questions
Q1. What does the %iowait value in CPU statistics mean?
A. %iowait represents the percentage of time that the CPU was idle during which the system had pending disk I/O requests. High iowait values indicate that the system is bottlenecked by disk operations.
Q2. How do I interpret the r_await and w_await columns?
A. r_await and w_await show the average time (in milliseconds) for read and write requests to be served, including the time spent in the queue and the service time. Higher values indicate slower I/O operations.
Q3. What's the difference between tps and IOPS?
A. tps (transfers per second) represents the number of I/O requests completed per second, regardless of the size. IOPS (I/O operations per second) is essentially the same metric but is often used in storage performance discussions.
Q4. How can I see statistics for a specific device only?
A. Use iostat -d device_name to show statistics only for the specified device.
References
https://man7.org/linux/man-pages/man1/iostat.1.html
Revisions
- 2025/05/05 First revision