iostat¶
Shows how busy the disks are: reads and writes per second, throughput, waiting times and utilization.

Syntax¶
iostat [OPTIONS] [DEVICE...] [INTERVAL [COUNT]]
Options¶
| Option | Description |
|---|---|
-x |
Extended statistics with waiting times and utilization |
-z |
Hide devices without activity |
-m, -h |
Values in MB, or human-readable |
-d |
Only disks, without the CPU summary |
-p sda |
Also show the partitions of a disk |
-y |
Skip the first report (averages since boot) |
The most important columns with -x:
| Column | Meaning |
|---|---|
r/s, w/s |
Read and write operations per second |
rkB/s, wkB/s |
Throughput |
r_await, w_await |
Average waiting time per operation in milliseconds |
aqu-sz |
Average length of the queue |
%util |
Share of the time the device was busy |
Examples¶
Throughput in MB since the boot:
iostat -dm
Linux 6.12.48+deb13-amd64 (tuxdex-srv) 10/26/2026 _x86_64_ (2 CPU)
Device tps MB_read/s MB_wrtn/s MB_dscd/s MB_read MB_wrtn MB_dscd
sda 4.18 0.04 0.21 0.00 42611 223904 0
sdb 1.02 0.06 0.18 0.00 61290 190341 0
sdc 0.84 0.01 0.05 0.00 10248 52867 0
Find the busy disk:
iostat -dxzy 2 1
Linux 6.12.48+deb13-amd64 (tuxdex-srv) 10/26/2026 _x86_64_ (2 CPU)
Device r/s rkB/s rrqm/s %rrqm r_await rareq-sz w/s wkB/s wrqm/s %wrqm w_await wareq-sz d/s dkB/s drqm/s %drqm d_await dareq-sz f/s f_await aqu-sz %util
sda 2.50 40.00 0.00 0.00 0.80 16.00 18.00 196.00 6.50 26.53 1.42 10.89 0.00 0.00 0.00 0.00 0.00 0.00 1.00 0.50 0.03 2.60
sdb 412.00 52736.00 0.00 0.00 11.84 128.00 188.50 24128.00 0.00 0.00 38.10 128.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 12.06 99.80
sdb (the backup disk under /mnt/backup) is busy all the time with long waiting times, sda with the system is idle. Here a backup job is running.
CPU summary with I/O wait:
iostat -c 2 2 | tail -3
avg-cpu: %user %nice %system %iowait %steal %idle
5.82 0.00 8.94 47.61 0.00 37.63
When do I need it?¶
- When
vmstatshows a highwaand you need to know which disk is busy - To check whether a disk or a storage is simply too slow (
await) - During backups or database maintenance, to see the load on the disks
%util on SSDs
SSDs and NVMe disks handle many requests in parallel. %util near 100 doesn't mean they
are at their limit; look at r_await and w_await instead.
Distribution differences¶
| Debian | Ubuntu | RHEL | Arch | |
|---|---|---|---|---|
| Install | sudo apt install sysstat |
sudo apt install sysstat |
sudo dnf install sysstat |
sudo pacman -S sysstat |