1BTT(1) BTT(1)
2
6 btt - analyse block i/o traces produces by blktrace
7
11 btt
12 [ -a | --seek-absolute ]
13 [ -A | --all-data ]
14 [ -B <output name> | --dump-blocknos=<output name> ]
15 [ -d <seconds> | --range-delta=<seconds> ]
16 [ -D <dev;...> | --devices=<dev;...> ]
17 [ -e <exe,...> | --exes=<exe,...> ]
18 [ -h | --help ]
19 [ -i <input name> | --input-file=<input name> ]
20 [ -I <output name> | --iostat=<output name> ]
21 [ -l <output name> | --d2c-latencies=<output name> ]
22 [ -L <freq> | --periodic-latencies=<freq> ]
23 [ -m <output name> | --seeks-per-second=<output name> ]
24 [ -M <dev map> | --dev-maps=<dev map>
25 [ -o <output name> | --output-file=<output name> ]
26 [ -p <output name> | --per-io-dump=<output name> ]
27 [ -P <output name> | --per-io-trees=<output name> ]
28 [ -q <output name> | --q2c-latencies=<output name> ]
29 [ -Q <output name> | --active-queue-depth=<output name> ]
30 [ -r | --no-remaps ]
31 [ -s <output name> | --seeks=<output name> ]
32 [ -S <interval> | --iostat-interval=<interval> ]
33 [ -t <sec> | --time-start=<sec> ]
34 [ -T <sec> | --time-end=<sec> ]
35 [ -u <output name> | --unplug-hist=<output name> ]
36 [ -v | --verbose ]
37 [ -V | --version ]
38 [ -X | --easy-parse-avgs ]
39 [ -z <output name> | --q2d-latencies=<output name> ]
40 [ -Z | --do-active ]
41
45 btt is a post-processing tool for the block layer IO tracing tool
46 called blktrace(8). As noted in its documentation, blktrace is a block
47 layer IO tracing mechanism which provides detailed information about
48 request queue operations up to user space.
49 btt will take in binary dump data from blkparse, and analyse the
51 events, producing a series of output from the analysis. It will also
52 build .dat files containing "range data" -- showing things like Q
53 activity (periods of time while Q events are being produced), C activ‐
54 ity (likewise for command completions), and etc.
55 Included with the distribution is a simple 3D plotting utility,
57 bno_plot, which can plot the block numbers btt outputs if the -B option
58 is specified. The display will display each IO generated, with the time
59 (seconds) along the X-axis, the block number (start) along the Y-axis
60 and the number of blocks transferred in the IO represented along the Z-
61 axis.
62
66 -a
67 --seek-absolute
68 When specified on the command line, this directs btt to calculate
69 seek distances based solely upon the ending block address of one
70 IO, and the start of the next. By default btt uses the concept of
71 the closeness to either the beginning or end of the previous IO.
72 See the Users Manual for more details about seek distances.
73 -A
75 --all-data
76 Normally btt will not print out verbose information concerning per-
77 process and per-device data. If you desire that level of detail
78 you can specify this option.
79 -B <output name>
81 --dump-blocknos=<output name>
82 This option will output absolute block numbers to three files pre‐
83 fixed by the specified output name:
84 prefix_device_r.dat
86 All read block numbers are output, first column is time
87 (seconds), second is the block number, and the third column
88 is the ending block number.
89 prefix_device_w.dat
91 All write block numbers are output, first column is time
92 (seconds), second is the block number, and the third column
93 is the ending block number.
94 prefix_device_c.dat
96 All block numbers (read and write) are output, first column
97 is time (seconds), second is the block number, and the third
98 column is the ending block number.
99 -d <seconds>
101 --range-delta=<seconds>
102 btt outputs a file containing Q and C activity, the notion of
103 active traces simply means that there are Q or C traces occurring
104 within a certain period of each other. The default values is 0.1
105 seconds; with this option allowing one to change that granularity.
106 The smaller the value, the more data points provided.
107 -D <dev;...>
109 --devices=<dev;...>
110 Normally, btt will produce data for all devices detected in the
111 traces parsed. With this option, one can reduce the analysis to one
112 or more devices provided in the string passed to this option. The
113 device identifiers are the major and minor number of each device,
114 and each device identifier is separated by a colon (:). A valid
115 specifier for devices 8,0 and 8,8 would then be: 8,0:8,8.
116 -e <exe,...>
118 --exes=<exe,...>
119 The -e option supplies the list of executables that will have I/Os
120 analysed.
121 -h
123 --help
124 Shows a short summary of possible command line option
125 -i <input name>
127 --input-file <input file>
128 Specifies the input file to analyse. This should be a trace file
129 produced by blktrace (8).
130 -I <output name>
132 --iostat=<output name>
133 The -I option directs btt to output iostat-like data to the speci‐
134 fied file. Refer to the iostat (sysstat) documentation for details
135 on the data columns.
136 -l <output name>
138 --d2c-latencies=<output name>
139 The -l option allows one to output per-IO D2C latencies respec‐
140 tively. The supplied argument provides the basis for the output
141 name for each device.
142 -L <freq>
144 --periodic-latencies=<freq>
145 The -L option allows one to output periodic latency information for
146 both Q2C and D2C latencies. The frequency specified will regulate
147 how often an average latency is output -- a floating point value
148 expressing seconds.
149 -m <output name>
151 --seeks-per-second=<output name>
152 Trigger btt to output seeks-per-second information. The first col‐
153 umn will contain a time value (seconds), and the second column
154 will indicate the number of seeks per second at that point.
155 -M <dev map>
157 --dev-maps=<dev map>
158 The -M option takes in a file generated by the provided script
159 (gen_disk_info.py), and allows for better output of device names.
160 -o <output name>
162 --output-file=<output name>
163 Specifies the output file name.
164 -p <output name>
166 --per-io-dump=<output name>
167 The -p option will generate a file that contains a list of all IO
168 "sequences" - showing the parts of each IO (Q, A, I/M, D, & C).
169 -P <output name>
171 --per-io-trees=<output name>
172 The -P option will generate a file that contains a list of all IO
173 "sequences" - showing only the Q, D & C operation times. The D & C
174 time values are separated from the Q time values with a vertical
175 bar.
176 -q <output name>
178 --q2c-latencies=<output name>
179 The -q option allows one to output per-IO Q2C latencies respec‐
180 tively. The supplied argument provides the basis for the output
181 name for each device.
182 -Q <output name>
184 --active-queue-depth=<output name>
185 The -Q option allows one to output data files showing the time
186 stamp and the depth of active commands (those issued but not com‐
187 pleted).
188 -r
190 --no-remaps
191 Ignore remap traces; older kernels did not implement the full remap
192 PDU.
193 -s <output name>
195 --seeks=<output name>
196 The -s option instructs btt to output seek data, the argument pro‐
197 vided is the basis for file names output. There are two files per
198 device, read seeks and write seeks.
199 -S <interval>
201 --iostat-interval=<interval>
202 The -S option specifies the interval to use between data output, it
203 defaults to once per second.
204 -t <sec>
206 --time-start=<sec>
207 -T <sec>
208 --time-end=<sec>
209 The -t/-T options allow one to set a start and/or end time for
210 analysing - analysing will only be done for traces after -t's argu‐
211 ment and before -T's argument. (-t and -T are optional, so if you
212 specify just -t, analysis will occur for all traces after the time
213 specified. Similarly, if only -T is specified, analysis stops after
214 -T's seconds.)
215 -u <output name>
217 --unplug-hist=<output name>
218 This option instructs btt to generate a data file containing his‐
219 togram information for unplug traces on a per device basis. It
220 shows how many times an unplug was hit with a specified number of
221 IOs released. There are 21 output values into the file, as follows:
222 a value of 0 represents 0..4 counts
224 a value of 1 represents 5..9 counts
225 a value of 2 represents 10..14 counts
226 etc, until
227 a value of 20 represents 100+ counts
228 The file name(s) generated use the text string passed as an argu‐
230 ment for the prefix, followed by the device identifier in
231 major,minor form, with a .dat extension. For example, with -u
232 up_hist specified on the command line: up_hist_008,032.dat.
233 -V
235 --version
236 Shows the version of btt.
237 -v
239 --verbose
240 Requests a more verbose output.
241 -X
243 --easy-parse-avgs
244 Provide data in an easy-to-parse form and write it to a file with
245 .avg exentsion
246 -z <output name>
248 --q2d-latencies=<output name>
249 The -z option allows one to output per-IO Q2D latencies respec‐
250 tively. The supplied argument provides the basis for the output
251 name for each device.
252 -Z
254 --do-active
255 The -Z will output files containing data which can be plotted show‐
256 ing per-device (and total system) I/O activity.
257
261 btt was written by Alan D. Brunelle. This man page was created from
262 the blktrace documentation by Bas Zoetekouw.
263
267 Report bugs to <linux-btrace@vger.kernel.org>
268
271 Copyright © 2006 Jens Axboe, Alan D. Brunelle and Nathan Scott.
272 This is free software. You may redistribute copies of it under the
273 terms of the GNU General Public License
274 <http://www.gnu.org/licenses/gpl.html>. There is NO WARRANTY, to the
275 extent permitted by law.
276 This manual page was created for Debian by Bas Zoetekouw. It was
277 derived from the documentation provided by the authors and it may be
278 used, distributed and modified under the terms of the GNU General Pub‐
279 lic License, version 2.
280 On Debian systems, the text of the GNU General Public License can be
281 found in /usr/share/common-licenses/GPL-2.
282
285 The btt Users Guide, which can be found in /usr/share/doc/blk‐
286 trace/btt.pdf
287 bno_plot (1), blktrace (8), blkparse (1), verify_blkparse (1),
288 blkrawverify (1), btt (1)
289blktrace git-20070910192508 September 29, 2007 BTT(1)