1biolatency(8) System Manager's Manual biolatency(8)
2
3
4
6 biolatency - Summarize block device I/O latency as a histogram.
7
9 biolatency [-h] [-F] [-T] [-Q] [-m] [-D] [-F] [-e] [-j] [-d DISK] [in‐
10 terval [count]]
11
13 biolatency traces block device I/O (disk I/O), and records the distri‐
14 bution of I/O latency (time). This is printed as a histogram either on
15 Ctrl-C, or after a given interval in seconds.
16
17 The latency of disk I/O operations is measured from when requests are
18 issued to the device up to completion. A -Q option can be used to in‐
19 clude time queued in the kernel.
20
21 This tool uses in-kernel eBPF maps for storing timestamps and the his‐
22 togram, for efficiency.
23
24 This works by tracing various kernel blk_*() functions using dynamic
25 tracing, and will need updating to match any changes to these func‐
26 tions.
27
28 Since this uses BPF, only the root user can use this tool.
29
31 CONFIG_BPF and bcc.
32
34 -h Print usage message.
35
36 -T Include timestamps on output.
37
38 -m Output histogram in milliseconds.
39
40 -D Print a histogram per disk device.
41
42 -F Print a histogram per set of I/O flags.
43
44 -j Print a histogram dictionary
45
46 -e Show extension summary(total, average)
47
48 -d DISK
49 Trace this disk only
50
51 interval
52 Output interval, in seconds.
53
54 count Number of outputs.
55
57 Summarize block device I/O latency as a histogram:
58 # biolatency
59
60 Print 1 second summaries, 10 times:
61 # biolatency 1 10
62
63 Print 1 second summaries, using milliseconds as units for the his‐
64 togram, and
65 include timestamps on output: # biolatency -mT 1
66
67 Include OS queued time in I/O time:
68 # biolatency -Q
69
70 Show a latency histogram for each disk device separately:
71 # biolatency -D
72
73 Show a latency histogram in a dictionary format:
74 # biolatency -j
75
76 Also show extension summary(total, average):
77 # biolatency -e
78
80 usecs Microsecond range
81
82 msecs Millisecond range
83
84 count How many I/O fell into this range
85
86 distribution
87 An ASCII bar chart to visualize the distribution (count column)
88
90 This traces kernel functions and maintains in-kernel timestamps and a
91 histogram, which are asynchronously copied to user-space. This method
92 is very efficient, and the overhead for most storage I/O rates (< 10k
93 IOPS) should be negligible. If you have a higher IOPS storage environ‐
94 ment, test and quantify the overhead before use.
95
97 This is from bcc.
98
99 https://github.com/iovisor/bcc
100
101 Also look in the bcc distribution for a companion _examples.txt file
102 containing example usage, output, and commentary for this tool.
103
105 Linux
106
108 Unstable - in development.
109
111 Brendan Gregg, Rocky Xing
112
114 biosnoop(8)
115
116
117
118USER COMMANDS 2020-12-30 biolatency(8)