ioping - Man Page

simple disk I/O latency monitoring tool

Examples (TL;DR)

Synopsys

ioping[-ABCDJLNRWGYykq] [-a count] [-b count] [-c count] [-e seed] [-i interval] [-l speed] [-r rate] [-t time] [-T time] [-s size] [-S wsize] [-o offset] [-w deadline] [-p period] [-P period] [-I [format]] directory|file|device
ioping-h | -v

Description

This tool generates various I/O patterns and lets you monitor I/O speed and latency in real time.

Options

-a,  -warmup count

Ignore in statistics first count requests, default 1. First request usually is much slower due to power-saving features.

-b,  -burst count

Make series of count requests without delay, default 0. Aggressive power-saving features slow down requests even after short delay.

-c,  -count count

Stop after count requests, default 0 (infinite).

-e,  -entropy seed

Set seed for random number generator, default 0 (random).

-i,  -interval time

Set time between requests, default 1s.

-l,  -speed-limit size

Set speed limit in size per second. Increases interval to request-size / speed.

-r,  -rate-limit count

Set rate limit in count per second. Increases interval to 1 / rate.

-t,  -min-time time

Minimal valid request time (0us). Too fast requests are ignored in statistics.

-T,  -max-time time

Maximum valid request time. Too slow requests are ignored in statistics.

-s,  -size size

Request size, default 4k.

-S,  -work-size size

Working set size (1m for directory, whole size for file or device).

-o,  -work-offset size

Starting offset in the file/device (0).

-w,  -work-time time

Stop after time passed, default 0 (infinite).

-p,  -print-count count

Print raw statistics for every count requests (see format below).

-P,  -print-interval time

Print raw statistics for every time.

-A,  -async

Use asynchronous I/O (io_setup(2), io_submit(2) etc syscalls).

-B,  -batch

Batch mode. Be quiet and print final statistics in raw format.

-C,  -cached

Use cached I/O. Suppress cache invalidation via posix_fadvise(2)) before read and fdatasync(2) after each write.

-D,  -direct

Use direct I/O (see O_DIRECT in open(2)).

-I,  -time [format]

Print current time for each request. Optional argument defines time format in strftime(3) notation, default is "%b %d %T" (Jan 01 00:00:00).

-J,  -json

Print output in JSON format.

-L,  -linear

Use sequential operations rather than random. This also sets default request size to 256k (as in -size 256k).

-N,  -nowait

Set RWF_NOWAIT on I/O, indicating to the kernel to do not wait if request cannot be executed immediately. (see RWF_NOWAIT in preadv2(2))

-R,  -rapid

Disk seek rate test, or bandwidth test if used together with -linear.

This option suppress human-readable output for each request (as -quiet), sets default interval to zero (-interval 0), stops measurement after 3 seconds (-work-time 3) and increases default working set size to 64m (-work-size 64m). Working set (-work-size) should be increased accordingly if disk has huge hardware cache.

-W,  -write

Use writes rather than reads. Safe for temporary file in directory target. Write I/O gives more reliable results for systems where non-cached reads are not supported or cached at some level.

Might be *DANGEROUS* for file/device: it will shred your data. In this case should be repeated three times (-WWW).

-G,  -read-write

Alternate read and write requests.

-Y,  -sync

Use sync I/O (see O_SYNC in open(2)).

-y,  -dsync

Use data sync I/O (see O_DSYNC in open(2)).

-k,  -keep

Keep and reuse temporary working file "ioping.tmp" (only for directory target).

-q,  -quiet

Suppress periodical human-readable output.

-h,  -help

Display help message and exit.

-v,  -version

Display version and exit.

Argument suffixes

For options that expect time argument (-interval, -print-interval and -work-time), default is seconds, unless you specify one of the following suffixes (case-insensitive):

nsnsec

nanoseconds (a billionth of a second, 1 / 1 000 000 000)

ususec

microseconds (a millionth of a second, 1 / 1 000 000)

msmsec

milliseconds (a thousandth of a second, 1 / 1 000)

ssec

seconds

mmin

minutes

hhour

hours

For options that expect "size" argument (-size, -speed-limit, -work-size and -work-offset), default is bytes, unless you specify one of the following suffixes (case-insensitive):

sector

disk sectors (a sector is always 512).

KiBkkb

kilobytes (1 024 bytes)

page

memory pages (a page is always 4KiB).

MiBmmb

megabytes (1 048 576 bytes)

GiBggb

gigabytes (1 073 741 824 bytes)

TiBttb

terabytes (1 099 511 627 776 bytes)

For options that expect "number" argument (-count and -print-count) you can optionally specify one of the following suffixes (case-insensitive):

k

kilo (thousands, 1 000)

m

mega (millions, 1 000 000)

g

giga (billions, 1 000 000 000)

t

tera (trillions, 1 000 000 000 000)

Exit Status

Returns 0 upon success. The following error codes are defined:

  1. Invalid usage (error in arguments).
  2. Error during preparation stage.
  3. Error during runtime.

Raw Statistics

ioping -print-count 100 -count 200 -interval 0 -quiet .
99 10970974 9024 36961531 90437 110818 358872 30756 100 12516420
100 9573265 10446 42785821 86849 95733 154609 10548 100 10649035
(1) (2)     (3)   (4)      (5)   (6)   (7)    (8)   (9) (10)

(1) count of requests in statistics
(2) running time         (nanoseconds)
(3) requests per second  (iops)
(4) transfer speed       (bytes per second)
(5) minimal request time (nanoseconds)
(6) average request time (nanoseconds)
(7) maximum request time (nanoseconds)
(8) request time standard deviation (nanoseconds)
(9) total requests       (including warmup, too slow or too fast)
(10) total running time  (nanoseconds)

JSON Output

With option -J|--json ioping prints json array of objects:
[
...
{
 // timestamps
 "timestamp": (unix time in seconds as float),
 "localtime": (local time ISO 8601),

 // io target
 "target": {
   "path": (target path),
   "fstype": (filesystem name),
   "device": (device name),
   "device_size": (device size in bytes)
 },

 // io request
 "io": {
   "request": (request index),
   "operation": (request type: "read" | "write"),
   "size": (request size in bytes),
   "time": (io time in ns),
   "ignored": (ignored in statistics: true | false)
 },

 // statistics
 "stat": {
   "count": (nr reqeusts),
   "size": (total io size in bytes),
   "time": (total io time in ns),
   "iops": (avg iops),
   "bps": (avg rate),
   "min": (min io time in ns),
   "avg": (avg io time in ns),
   "max": (max io time in ns),
   "mdev": (standard deviation in ns)
 },

 // load statistics
 "load": {
   "count": (nr requests),
   "size": (total io size in bytes),
   "time": (total real time in ns),
   "iops": (avg iops),
   "bps": (avg rate)
 },
},
...
]

Examples

ioping .

Show disk I/O latency using the default values and the current directory, until interrupted. This command prepares temporary (unlinked/hidden) working file and reads random chunks from it using non-cached read requests.

ioping -c 10 -s 1M /tmp

Measure latency on /tmp using 10 requests of 1 megabyte each.

ioping -R /dev/sda

Measure disk seek rate.

ioping -RL /dev/sda

Measure disk sequential speed.

ioping -RLB . | awk '{print $4}'

Get disk sequential speed in bytes per second.

ioping -J . | jq -r --stream 'fromstream(1|truncate_stream(inputs)) | [.localtime, .io.time/1000000] | @tsv'

Select localtime and io time in milliseconds from json outout.

See Also

iostat(1), dd(1), fio(1), stress(1), stress-ng(1), dbench(1), sysbench(1), fsstress, xfstests, hdparm(8), badblocks(8),

Homepage

https://github.com/koct9i/ioping/.

Authors

This program was written by Konstantin Khlebnikov koct9i@gmail.com.
Man-page was written by Kir Kolyshkin kir@openvz.org.

Info

Oct 2014