1 ===================== 1 =====================
2 I/O statistics fields 2 I/O statistics fields
3 ===================== 3 =====================
4 4
5 Since 2.4.20 (and some versions before, with p 5 Since 2.4.20 (and some versions before, with patches), and 2.5.45,
6 more extensive disk statistics have been intro 6 more extensive disk statistics have been introduced to help measure disk
7 activity. Tools such as ``sar`` and ``iostat`` 7 activity. Tools such as ``sar`` and ``iostat`` typically interpret these and do
8 the work for you, but in case you are interest 8 the work for you, but in case you are interested in creating your own
9 tools, the fields are explained here. 9 tools, the fields are explained here.
10 10
11 In 2.4 now, the information is found as additi 11 In 2.4 now, the information is found as additional fields in
12 ``/proc/partitions``. In 2.6 and upper, the s 12 ``/proc/partitions``. In 2.6 and upper, the same information is found in two
13 places: one is in the file ``/proc/diskstats`` 13 places: one is in the file ``/proc/diskstats``, and the other is within
14 the sysfs file system, which must be mounted i 14 the sysfs file system, which must be mounted in order to obtain
15 the information. Throughout this document we'l 15 the information. Throughout this document we'll assume that sysfs
16 is mounted on ``/sys``, although of course it 16 is mounted on ``/sys``, although of course it may be mounted anywhere.
17 Both ``/proc/diskstats`` and sysfs use the sam 17 Both ``/proc/diskstats`` and sysfs use the same source for the information
18 and so should not differ. 18 and so should not differ.
19 19
20 Here are examples of these different formats:: 20 Here are examples of these different formats::
21 21
22 2.4: 22 2.4:
23 3 0 39082680 hda 446216 784926 955 23 3 0 39082680 hda 446216 784926 9550688 4382310 424847 312726 5922052 19310380 0 3376340 23705160
24 3 1 9221278 hda1 35486 0 35496 38 24 3 1 9221278 hda1 35486 0 35496 38030 0 0 0 0 0 38030 38030
25 25
26 2.6+ sysfs: 26 2.6+ sysfs:
27 446216 784926 9550688 4382310 424847 312 27 446216 784926 9550688 4382310 424847 312726 5922052 19310380 0 3376340 23705160
28 35486 38030 38030 38030 28 35486 38030 38030 38030
29 29
30 2.6+ diskstats: 30 2.6+ diskstats:
31 3 0 hda 446216 784926 9550688 43823 31 3 0 hda 446216 784926 9550688 4382310 424847 312726 5922052 19310380 0 3376340 23705160
32 3 1 hda1 35486 38030 38030 38030 32 3 1 hda1 35486 38030 38030 38030
33 33
34 4.18+ diskstats: 34 4.18+ diskstats:
35 3 0 hda 446216 784926 9550688 43823 35 3 0 hda 446216 784926 9550688 4382310 424847 312726 5922052 19310380 0 3376340 23705160 0 0 0 0
36 36
37 On 2.4 you might execute ``grep 'hda ' /proc/p 37 On 2.4 you might execute ``grep 'hda ' /proc/partitions``. On 2.6+, you have
38 a choice of ``cat /sys/block/hda/stat`` or ``g 38 a choice of ``cat /sys/block/hda/stat`` or ``grep 'hda ' /proc/diskstats``.
39 39
40 The advantage of one over the other is that th 40 The advantage of one over the other is that the sysfs choice works well
41 if you are watching a known, small set of disk 41 if you are watching a known, small set of disks. ``/proc/diskstats`` may
42 be a better choice if you are watching a large 42 be a better choice if you are watching a large number of disks because
43 you'll avoid the overhead of 50, 100, or 500 o 43 you'll avoid the overhead of 50, 100, or 500 or more opens/closes with
44 each snapshot of your disk statistics. 44 each snapshot of your disk statistics.
45 45
46 In 2.4, the statistics fields are those after 46 In 2.4, the statistics fields are those after the device name. In
47 the above example, the first field of statisti 47 the above example, the first field of statistics would be 446216.
48 By contrast, in 2.6+ if you look at ``/sys/blo 48 By contrast, in 2.6+ if you look at ``/sys/block/hda/stat``, you'll
49 find just the 15 fields, beginning with 446216 49 find just the 15 fields, beginning with 446216. If you look at
50 ``/proc/diskstats``, the 15 fields will be pre 50 ``/proc/diskstats``, the 15 fields will be preceded by the major and
51 minor device numbers, and device name. Each o 51 minor device numbers, and device name. Each of these formats provides
52 15 fields of statistics, each meaning exactly 52 15 fields of statistics, each meaning exactly the same things.
53 All fields except field 9 are cumulative since 53 All fields except field 9 are cumulative since boot. Field 9 should
54 go to zero as I/Os complete; all others only i 54 go to zero as I/Os complete; all others only increase (unless they
55 overflow and wrap). Wrapping might eventually 55 overflow and wrap). Wrapping might eventually occur on a very busy
56 or long-lived system; so applications should b 56 or long-lived system; so applications should be prepared to deal with
57 it. Regarding wrapping, the types of the field 57 it. Regarding wrapping, the types of the fields are either unsigned
58 int (32 bit) or unsigned long (32-bit or 64-bi 58 int (32 bit) or unsigned long (32-bit or 64-bit, depending on your
59 machine) as noted per-field below. Unless your 59 machine) as noted per-field below. Unless your observations are very
60 spread in time, these fields should not wrap t 60 spread in time, these fields should not wrap twice before you notice it.
61 61
62 Each set of stats only applies to the indicate 62 Each set of stats only applies to the indicated device; if you want
63 system-wide stats you'll have to find all the 63 system-wide stats you'll have to find all the devices and sum them all up.
64 64
65 Field 1 -- # of reads completed (unsigned lon 65 Field 1 -- # of reads completed (unsigned long)
66 This is the total number of reads complete 66 This is the total number of reads completed successfully.
67 67
68 Field 2 -- # of reads merged, field 6 -- # of 68 Field 2 -- # of reads merged, field 6 -- # of writes merged (unsigned long)
69 Reads and writes which are adjacent to eac 69 Reads and writes which are adjacent to each other may be merged for
70 efficiency. Thus two 4K reads may become 70 efficiency. Thus two 4K reads may become one 8K read before it is
71 ultimately handed to the disk, and so it w 71 ultimately handed to the disk, and so it will be counted (and queued)
72 as only one I/O. This field lets you know 72 as only one I/O. This field lets you know how often this was done.
73 73
74 Field 3 -- # of sectors read (unsigned long) 74 Field 3 -- # of sectors read (unsigned long)
75 This is the total number of sectors read s 75 This is the total number of sectors read successfully.
76 76
77 Field 4 -- # of milliseconds spent reading (u 77 Field 4 -- # of milliseconds spent reading (unsigned int)
78 This is the total number of milliseconds s 78 This is the total number of milliseconds spent by all reads (as
79 measured from blk_mq_alloc_request() to __ !! 79 measured from __make_request() to end_that_request_last()).
80 80
81 Field 5 -- # of writes completed (unsigned lo 81 Field 5 -- # of writes completed (unsigned long)
82 This is the total number of writes complet 82 This is the total number of writes completed successfully.
83 83
84 Field 6 -- # of writes merged (unsigned long 84 Field 6 -- # of writes merged (unsigned long)
85 See the description of field 2. 85 See the description of field 2.
86 86
87 Field 7 -- # of sectors written (unsigned lon 87 Field 7 -- # of sectors written (unsigned long)
88 This is the total number of sectors writte 88 This is the total number of sectors written successfully.
89 89
90 Field 8 -- # of milliseconds spent writing (u 90 Field 8 -- # of milliseconds spent writing (unsigned int)
91 This is the total number of milliseconds s 91 This is the total number of milliseconds spent by all writes (as
92 measured from blk_mq_alloc_request() to __ !! 92 measured from __make_request() to end_that_request_last()).
93 93
94 Field 9 -- # of I/Os currently in progress (u 94 Field 9 -- # of I/Os currently in progress (unsigned int)
95 The only field that should go to zero. Inc 95 The only field that should go to zero. Incremented as requests are
96 given to appropriate struct request_queue 96 given to appropriate struct request_queue and decremented as they finish.
97 97
98 Field 10 -- # of milliseconds spent doing I/Os 98 Field 10 -- # of milliseconds spent doing I/Os (unsigned int)
99 This field increases so long as field 9 is 99 This field increases so long as field 9 is nonzero.
100 100
101 Since 5.0 this field counts jiffies when a 101 Since 5.0 this field counts jiffies when at least one request was
102 started or completed. If request runs more 102 started or completed. If request runs more than 2 jiffies then some
103 I/O time might be not accounted in case of 103 I/O time might be not accounted in case of concurrent requests.
104 104
105 Field 11 -- weighted # of milliseconds spent d 105 Field 11 -- weighted # of milliseconds spent doing I/Os (unsigned int)
106 This field is incremented at each I/O star 106 This field is incremented at each I/O start, I/O completion, I/O
107 merge, or read of these stats by the numbe 107 merge, or read of these stats by the number of I/Os in progress
108 (field 9) times the number of milliseconds 108 (field 9) times the number of milliseconds spent doing I/O since the
109 last update of this field. This can provi 109 last update of this field. This can provide an easy measure of both
110 I/O completion time and the backlog that m 110 I/O completion time and the backlog that may be accumulating.
111 111
112 Field 12 -- # of discards completed (unsigned 112 Field 12 -- # of discards completed (unsigned long)
113 This is the total number of discards compl 113 This is the total number of discards completed successfully.
114 114
115 Field 13 -- # of discards merged (unsigned lon 115 Field 13 -- # of discards merged (unsigned long)
116 See the description of field 2 116 See the description of field 2
117 117
118 Field 14 -- # of sectors discarded (unsigned l 118 Field 14 -- # of sectors discarded (unsigned long)
119 This is the total number of sectors discar 119 This is the total number of sectors discarded successfully.
120 120
121 Field 15 -- # of milliseconds spent discarding 121 Field 15 -- # of milliseconds spent discarding (unsigned int)
122 This is the total number of milliseconds s 122 This is the total number of milliseconds spent by all discards (as
123 measured from blk_mq_alloc_request() to __ !! 123 measured from __make_request() to end_that_request_last()).
124 124
125 Field 16 -- # of flush requests completed 125 Field 16 -- # of flush requests completed
126 This is the total number of flush requests 126 This is the total number of flush requests completed successfully.
127 127
128 Block layer combines flush requests and ex 128 Block layer combines flush requests and executes at most one at a time.
129 This counts flush requests executed by dis 129 This counts flush requests executed by disk. Not tracked for partitions.
130 130
131 Field 17 -- # of milliseconds spent flushing 131 Field 17 -- # of milliseconds spent flushing
132 This is the total number of milliseconds s 132 This is the total number of milliseconds spent by all flush requests.
133 133
134 To avoid introducing performance bottlenecks, 134 To avoid introducing performance bottlenecks, no locks are held while
135 modifying these counters. This implies that m 135 modifying these counters. This implies that minor inaccuracies may be
136 introduced when changes collide, so (for insta 136 introduced when changes collide, so (for instance) adding up all the
137 read I/Os issued per partition should equal th 137 read I/Os issued per partition should equal those made to the disks ...
138 but due to the lack of locking it may only be 138 but due to the lack of locking it may only be very close.
139 139
140 In 2.6+, there are counters for each CPU, whic 140 In 2.6+, there are counters for each CPU, which make the lack of locking
141 almost a non-issue. When the statistics are r 141 almost a non-issue. When the statistics are read, the per-CPU counters
142 are summed (possibly overflowing the unsigned 142 are summed (possibly overflowing the unsigned long variable they are
143 summed to) and the result given to the user. 143 summed to) and the result given to the user. There is no convenient
144 user interface for accessing the per-CPU count 144 user interface for accessing the per-CPU counters themselves.
145 145
146 Since 4.19 request times are measured with nan 146 Since 4.19 request times are measured with nanoseconds precision and
147 truncated to milliseconds before showing in th 147 truncated to milliseconds before showing in this interface.
148 148
149 Disks vs Partitions 149 Disks vs Partitions
150 ------------------- 150 -------------------
151 151
152 There were significant changes between 2.4 and 152 There were significant changes between 2.4 and 2.6+ in the I/O subsystem.
153 As a result, some statistic information disapp 153 As a result, some statistic information disappeared. The translation from
154 a disk address relative to a partition to the 154 a disk address relative to a partition to the disk address relative to
155 the host disk happens much earlier. All merge 155 the host disk happens much earlier. All merges and timings now happen
156 at the disk level rather than at both the disk 156 at the disk level rather than at both the disk and partition level as
157 in 2.4. Consequently, you'll see a different 157 in 2.4. Consequently, you'll see a different statistics output on 2.6+ for
158 partitions from that for disks. There are onl 158 partitions from that for disks. There are only *four* fields available
159 for partitions on 2.6+ machines. This is refl 159 for partitions on 2.6+ machines. This is reflected in the examples above.
160 160
161 Field 1 -- # of reads issued 161 Field 1 -- # of reads issued
162 This is the total number of reads issued t 162 This is the total number of reads issued to this partition.
163 163
164 Field 2 -- # of sectors read 164 Field 2 -- # of sectors read
165 This is the total number of sectors reques 165 This is the total number of sectors requested to be read from this
166 partition. 166 partition.
167 167
168 Field 3 -- # of writes issued 168 Field 3 -- # of writes issued
169 This is the total number of writes issued 169 This is the total number of writes issued to this partition.
170 170
171 Field 4 -- # of sectors written 171 Field 4 -- # of sectors written
172 This is the total number of sectors reques 172 This is the total number of sectors requested to be written to
173 this partition. 173 this partition.
174 174
175 Note that since the address is translated to a 175 Note that since the address is translated to a disk-relative one, and no
176 record of the partition-relative address is ke 176 record of the partition-relative address is kept, the subsequent success
177 or failure of the read cannot be attributed to 177 or failure of the read cannot be attributed to the partition. In other
178 words, the number of reads for partitions is c 178 words, the number of reads for partitions is counted slightly before time
179 of queuing for partitions, and at completion f 179 of queuing for partitions, and at completion for whole disks. This is
180 a subtle distinction that is probably unintere 180 a subtle distinction that is probably uninteresting for most cases.
181 181
182 More significant is the error induced by count 182 More significant is the error induced by counting the numbers of
183 reads/writes before merges for partitions and 183 reads/writes before merges for partitions and after for disks. Since a
184 typical workload usually contains a lot of suc 184 typical workload usually contains a lot of successive and adjacent requests,
185 the number of reads/writes issued can be sever 185 the number of reads/writes issued can be several times higher than the
186 number of reads/writes completed. 186 number of reads/writes completed.
187 187
188 In 2.6.25, the full statistic set is again ava 188 In 2.6.25, the full statistic set is again available for partitions and
189 disk and partition statistics are consistent a 189 disk and partition statistics are consistent again. Since we still don't
190 keep record of the partition-relative address, 190 keep record of the partition-relative address, an operation is attributed to
191 the partition which contains the first sector 191 the partition which contains the first sector of the request after the
192 eventual merges. As requests can be merged acr 192 eventual merges. As requests can be merged across partition, this could lead
193 to some (probably insignificant) inaccuracy. 193 to some (probably insignificant) inaccuracy.
194 194
195 Additional notes 195 Additional notes
196 ---------------- 196 ----------------
197 197
198 In 2.6+, sysfs is not mounted by default. If 198 In 2.6+, sysfs is not mounted by default. If your distribution of
199 Linux hasn't added it already, here's the line 199 Linux hasn't added it already, here's the line you'll want to add to
200 your ``/etc/fstab``:: 200 your ``/etc/fstab``::
201 201
202 none /sys sysfs defaults 0 0 202 none /sys sysfs defaults 0 0
203 203
204 204
205 In 2.6+, all disk statistics were removed from 205 In 2.6+, all disk statistics were removed from ``/proc/stat``. In 2.4, they
206 appear in both ``/proc/partitions`` and ``/pro 206 appear in both ``/proc/partitions`` and ``/proc/stat``, although the ones in
207 ``/proc/stat`` take a very different format fr 207 ``/proc/stat`` take a very different format from those in ``/proc/partitions``
208 (see proc(5), if your system has it.) 208 (see proc(5), if your system has it.)
209 209
210 -- ricklind@us.ibm.com 210 -- ricklind@us.ibm.com
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.