blob: 0dbc946de2eab8f2c326e48f592d364fd6196d26 [file] [log] [blame]
Andy Isaacson37a32792006-01-08 01:04:00 -08001Block layer statistics in /sys/block/<dev>/stat
2===============================================
3
4This file documents the contents of the /sys/block/<dev>/stat file.
5
6The stat file provides several statistics about the state of block
7device <dev>.
8
9Q. Why are there multiple statistics in a single file? Doesn't sysfs
10 normally contain a single value per file?
11A. By having a single file, the kernel can guarantee that the statistics
12 represent a consistent snapshot of the state of the device. If the
13 statistics were exported as multiple files containing one statistic
14 each, it would be impossible to guarantee that a set of readings
15 represent a single point in time.
16
17The stat file consists of a single line of text containing 11 decimal
18values separated by whitespace. The fields are summarized in the
19following table, and described in more detail below.
20
21Name units description
22---- ----- -----------
23read I/Os requests number of read I/Os processed
24read merges requests number of read I/Os merged with in-queue I/O
25read sectors sectors number of sectors read
26read ticks milliseconds total wait time for read requests
27write I/Os requests number of write I/Os processed
28write merges requests number of write I/Os merged with in-queue I/O
29write sectors sectors number of sectors written
30write ticks milliseconds total wait time for write requests
31in_flight requests number of I/Os currently in flight
32io_ticks milliseconds total time this block device has been active
33time_in_queue milliseconds total wait time for all requests
34
35read I/Os, write I/Os
36=====================
37
38These values increment when an I/O request completes.
39
40read merges, write merges
41=========================
42
43These values increment when an I/O request is merged with an
44already-queued I/O request.
45
46read sectors, write sectors
47===========================
48
49These values count the number of sectors read from or written to this
50block device. The "sectors" in question are the standard UNIX 512-byte
51sectors, not any device- or filesystem-specific block size. The
52counters are incremented when the I/O completes.
53
54read ticks, write ticks
55=======================
56
57These values count the number of milliseconds that I/O requests have
58waited on this block device. If there are multiple I/O requests waiting,
59these values will increase at a rate greater than 1000/second; for
60example, if 60 read requests wait for an average of 30 ms, the read_ticks
61field will increase by 60*30 = 1800.
62
63in_flight
64=========
65
66This value counts the number of I/O requests that have been issued to
67the device driver but have not yet completed. It does not include I/O
68requests that are in the queue but not yet issued to the device driver.
69
70io_ticks
71========
72
73This value counts the number of milliseconds during which the device has
74had I/O requests queued.
75
76time_in_queue
77=============
78
79This value counts the number of milliseconds that I/O requests have waited
80on this block device. If there are multiple I/O requests waiting, this
81value will increase as the product of the number of milliseconds times the
82number of requests waiting (see "read ticks" above for an example).