xref: /freebsd/usr.sbin/iostat/iostat.8 (revision edf8578117e8844e02c0121147f45e4609b30680)
1.\"
2.\" Copyright (c) 1997 Kenneth D. Merry.
3.\" All rights reserved.
4.\"
5.\" Redistribution and use in source and binary forms, with or without
6.\" modification, are permitted provided that the following conditions
7.\" are met:
8.\" 1. Redistributions of source code must retain the above copyright
9.\"    notice, this list of conditions and the following disclaimer.
10.\" 2. Redistributions in binary form must reproduce the above copyright
11.\"    notice, this list of conditions and the following disclaimer in the
12.\"    documentation and/or other materials provided with the distribution.
13.\" 3. The name of the author may not be used to endorse or promote products
14.\"    derived from this software without specific prior written permission.
15.\"
16.\" THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
17.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
18.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
19.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
20.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
21.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
22.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
23.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
24.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
25.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
26.\" SUCH DAMAGE.
27.\"
28.\" Copyright (c) 1985, 1991, 1993
29.\"	The Regents of the University of California.  All rights reserved.
30.\"
31.\" Redistribution and use in source and binary forms, with or without
32.\" modification, are permitted provided that the following conditions
33.\" are met:
34.\" 1. Redistributions of source code must retain the above copyright
35.\"    notice, this list of conditions and the following disclaimer.
36.\" 2. Redistributions in binary form must reproduce the above copyright
37.\"    notice, this list of conditions and the following disclaimer in the
38.\"    documentation and/or other materials provided with the distribution.
39.\" 3. Neither the name of the University nor the names of its contributors
40.\"    may be used to endorse or promote products derived from this software
41.\"    without specific prior written permission.
42.\"
43.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
44.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
45.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
46.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
47.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
48.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
49.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
50.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
51.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
52.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
53.\" SUCH DAMAGE.
54.\"
55.\"	@(#)iostat.8	8.1 (Berkeley) 6/6/93
56.\"
57.Dd August 29, 2023
58.Dt IOSTAT 8
59.Os
60.Sh NAME
61.Nm iostat
62.Nd report
63.Tn I/O
64statistics
65.Sh SYNOPSIS
66.Nm
67.Op Fl CdhIKoTxz
68.Op Fl c Ar count
69.Op Fl M Ar core
70.Op Fl n Ar devs
71.Op Fl N Ar system
72.Oo
73.Fl t
74.Sm off
75.Ar type , if , pass
76.Sm on
77.Oc
78.Op Fl w Ar wait
79.Op Ar drives
80.Sh DESCRIPTION
81The
82.Nm
83utility displays kernel
84.Tn I/O
85statistics on terminal, device and cpu operations.
86The first statistics that are printed are averaged over the system uptime.
87To get information about the current activity, a suitable wait time should
88be specified, so that the subsequent sets of printed statistics will be
89averaged over that time.
90.Pp
91The options are as follows:
92.Bl -tag -width flag
93.It Fl C
94Display CPU statistics.
95This is on by default, unless
96.Fl d
97or
98.Fl x
99is specified.
100.It Fl c
101Repeat the display
102.Ar count
103times.
104If no repeat
105.Ar count
106is specified, the default depends on whether
107.Fl w
108is specified.
109With
110.Fl w
111the default repeat count is infinity, otherwise it is 1.
112.It Fl d
113Display only device statistics.
114If this flag is turned on, only device statistics will be displayed, unless
115.Fl C
116or
117.Fl T
118is also specified to enable the display of CPU or TTY statistics.
119.It Fl h
120Put
121.Nm
122in
123.Sq top
124mode.
125In this mode,
126.Nm
127will show devices in order from highest to lowest bytes
128per measurement cycle.
129.It Fl I
130Display total statistics for a given time period, rather than average
131statistics for each second during that time period.
132.It Fl K
133In the blocks transferred display (-o), display block count in kilobytes rather
134then the device native block size.
135.It Fl M
136Extract values associated with the name list from the specified core
137instead of the default
138.Dq Pa /dev/kmem .
139.It Fl N
140Extract the name list from the specified system instead of the default
141.Dq Pa /boot/kernel/kernel .
142.It Fl n
143Display up to
144.Ar devs
145number of devices.
146The
147.Nm
148utility will display fewer devices if there are not
149.Ar devs
150devices present.
151.It Fl o
152Display old-style
153.Nm
154device statistics.
155Sectors per second, transfers per second, and milliseconds per seek are
156displayed.
157If
158.Fl I
159is specified, total blocks/sectors, total transfers, and
160milliseconds per seek are displayed.
161.It Fl T
162Display TTY statistics.
163This is on by default, unless
164.Fl d
165or
166.Fl x
167is specified.
168.It Fl t
169Specify which types of devices to display.
170There are three different categories of devices:
171.Pp
172.Bl -tag -width indent -compact
173.It device type:
174.Bl -tag -width 9n -compact
175.It da
176Direct Access devices
177.It sa
178Sequential Access devices
179.It printer
180Printers
181.It proc
182Processor devices
183.It worm
184Write Once Read Multiple devices
185.It cd
186CD devices
187.It scanner
188Scanner devices
189.It optical
190Optical Memory devices
191.It changer
192Medium Changer devices
193.It comm
194Communication devices
195.It array
196Storage Array devices
197.It enclosure
198Enclosure Services devices
199.It floppy
200Floppy devices
201.El
202.Pp
203.It interface:
204.Bl -tag -width 9n -compact
205.It IDE
206Integrated Drive Electronics devices
207.It SCSI
208Small Computer System Interface devices
209.It other
210Any other device interface
211.El
212.Pp
213.It passthrough:
214.Bl -tag -width 9n -compact
215.It pass
216Passthrough devices
217.El
218.El
219.Pp
220The user must specify at least one device type, and may specify at most
221one device type from each category.
222Multiple device types in a single device type statement must be separated by
223commas.
224.Pp
225Any number of
226.Fl t
227arguments may be specified on the command line.
228All
229.Fl t
230arguments are ORed together to form a matching expression against which
231all devices in the system are compared.
232Any device that fully matches any
233.Fl t
234argument will be included in the
235.Nm
236output, up to the number of devices that can be displayed in
23780 columns, or the maximum number of devices specified by the user.
238.It Fl w
239Pause
240.Ar wait
241seconds between each display.
242If no
243.Ar wait
244interval is specified, the default is 1 second.
245.Pp
246The
247.Nm
248command will accept and honor a non-integer number of seconds.
249Note that the interval only has millisecond granularity.
250Finer values will be truncated.
251E.g.,
252.Dq Li -w1.0001
253is the same as
254.Dq Li -w1.000 .
255The interval will also suffer from modifications to
256.Va kern.hz
257so your mileage may vary.
258.It Fl x
259Show extended disk statistics.
260Each disk is displayed on a line of its own with all available statistics.
261If this flag is turned on, only disk statistics will be displayed, unless
262.Fl C
263or
264.Fl T
265is also specified to enable the display of CPU or TTY statistics.
266.It Fl z
267If
268.Fl x
269is specified, omit lines for devices with no activity.
270.El
271.Pp
272The
273.Nm
274utility displays its information in the following format:
275.Bl -tag -width flag
276.It tty
277.Bl -tag -width indent -compact
278.It tin
279characters read from terminals
280.It tout
281characters written to terminals
282.El
283.It devices
284Device operations.
285The header of the field is the device name and unit number.
286The
287.Nm
288utility
289will display as many devices as will fit in a standard 80 column screen, or
290the maximum number of devices in the system, whichever is smaller.
291If
292.Fl n
293is specified on the command line,
294.Nm
295will display the smaller of the
296requested number of devices, and the maximum number of devices in the system.
297To force
298.Nm
299to display specific drives, their names may be supplied on the command
300line.
301The
302.Nm
303utility
304will not display more devices than will fit in an 80 column screen, unless
305the
306.Fl n
307argument is given on the command line to specify a maximum number of
308devices to display.
309If fewer devices are specified on the command line than will fit in an 80
310column screen,
311.Nm
312will show only the specified devices.
313.Pp
314The standard
315.Nm
316device display shows the following statistics:
317.Pp
318.Bl -tag -width indent -compact
319.It KB/t
320kilobytes per transfer
321.It tps
322transfers per second
323.It MB/s
324megabytes per second
325.El
326.Pp
327The standard
328.Nm
329device display, with the
330.Fl I
331flag specified, shows the following statistics:
332.Pp
333.Bl -tag -width indent -compact
334.It KB/t
335kilobytes per transfer
336.It xfrs
337total number of transfers
338.It MB
339total number of megabytes transferred
340.El
341.Pp
342The extended
343.Nm
344device display, with the
345.Fl x
346flag specified, shows the following statistics:
347.Pp
348.Bl -tag -width indent -compact
349.It r/s
350read operations per second
351.It w/s
352write operations per second
353.It kr/s
354kilobytes read per second
355.It kw/s
356kilobytes write per second
357.It qlen
358transactions queue length
359.It ms/r
360average duration of read transactions, in milliseconds
361.It ms/w
362average duration of write transactions, in milliseconds
363.It ms/o
364average duration of all other transactions, in milliseconds
365.It ms/t
366average duration of all transactions, in milliseconds
367.It %b
368% of time the device had one or more outstanding transactions
369.El
370.Pp
371The extended
372.Nm
373device display, with the
374.Fl x
375and
376.Fl I
377flags specified, shows the following statistics:
378.Pp
379.Bl -tag -width indent -compact
380.It r/i
381read operations per time period
382.It w/i
383write operations per time period
384.It kr/i
385kilobytes read per time period
386.It kw/i
387kilobytes write per time period
388.It qlen
389transactions queue length
390.It tsvc_t/i
391total duration of transactions per time period, in seconds
392.It sb/i
393total time the device had one or more outstanding transactions per
394time period, in seconds
395.El
396.Pp
397The old-style
398.Nm
399display (using
400.Fl o )
401shows the following statistics:
402.Pp
403.Bl -tag -width indent -compact
404.It sps
405sectors transferred per second
406.It tps
407transfers per second
408.It msps
409average milliseconds per transaction
410.El
411.Pp
412The old-style
413.Nm
414display, with the
415.Fl I
416flag specified, shows the following statistics:
417.Pp
418.Bl -tag -width indent -compact
419.It blk
420total blocks/sectors transferred
421.It xfr
422total transfers
423.It msps
424average milliseconds per transaction
425.El
426.It cpu
427.Bl -tag -width indent -compact
428.It \&us
429% of cpu time in user mode
430.It \&ni
431% of cpu time in user mode running niced processes
432.It \&sy
433% of cpu time in system mode
434.It \&in
435% of cpu time in interrupt mode
436.It \&id
437% of cpu time in idle mode
438.El
439.El
440.Sh FILES
441.Bl -tag -width /boot/kernel/kernel -compact
442.It Pa /boot/kernel/kernel
443Default kernel namelist.
444.It Pa /dev/kmem
445Default memory file.
446.El
447.Sh EXAMPLES
448.Dl iostat -w 1 da0 da1 cd0
449.Pp
450Display statistics for the first two Direct Access devices and the first
451CDROM device every second ad infinitum.
452.Pp
453.Dl iostat -c 2
454.Pp
455Display the statistics for the first four devices in the system twice, with
456a one second display interval.
457.Pp
458.Dl iostat -t da -t cd -w 1
459.Pp
460Display statistics for all CDROM and Direct Access devices every second
461ad infinitum.
462.Pp
463.Dl iostat -t da,scsi,pass -t cd,scsi,pass
464.Pp
465Display statistics once for all SCSI passthrough devices that provide access
466to either Direct Access or CDROM devices.
467.Pp
468.Dl iostat -h -n 8 -w 1
469.Pp
470Display up to 8 devices with the most I/O every second ad infinitum.
471.Pp
472.Dl iostat -dh -t da -w 1
473.Pp
474Omit the TTY and CPU displays, show devices in order of performance and
475show only Direct Access devices every second ad infinitum.
476.Pp
477.Dl iostat -Iw 3
478.Pp
479Display total statistics every three seconds ad infinitum.
480.Pp
481.Dl iostat -odICTw 2 -c 9
482.Pp
483Display total statistics using the old-style output format 9 times, with
484a two second interval between each measurement/display.
485The
486.Fl d
487flag generally disables the TTY and CPU displays, but since the
488.Fl T
489and
490.Fl C
491flags are given, the TTY and CPU displays will be displayed.
492.Sh SEE ALSO
493.Xr fstat 1 ,
494.Xr netstat 1 ,
495.Xr nfsstat 1 ,
496.Xr ps 1 ,
497.Xr systat 1 ,
498.Xr devstat 3 ,
499.Xr ctlstat 8 ,
500.Xr gstat 8 ,
501.Xr pstat 8 ,
502.Xr vmstat 8
503.Pp
504The sections starting with ``Interpreting system activity'' in
505.%T "Installing and Operating 4.3BSD" .
506.Sh HISTORY
507This version of
508.Nm
509first appeared in
510.Fx 3.0 .
511.Sh AUTHORS
512.An Kenneth Merry Aq Mt ken@FreeBSD.org
513.Sh BUGS
514The use of
515.Nm
516as a debugging tool for crash dumps is probably limited because there is
517currently no way to get statistics that only cover the time immediately before
518the crash.
519