table of contents
| BTT(1) | BTT(1) |
NAME¶
btt - analyse block i/o traces produces by blktraceSYNOPSIS¶
bttDESCRIPTION¶
btt is a post-processing tool for the block layer IO tracing tool called blktrace(8). As noted in its documentation, blktrace is a block layer IO tracing mechanism which provides detailed information about request queue operations up to user space.OPTIONS¶
-aWhen specified on the command line, this
directs btt to calculate seek distances based solely upon the ending block
address of one IO, and the start of the next. By default btt uses the
concept of the closeness to either the beginning or end of the previous IO.
See the Users Manual for more details about seek distances.
Normally btt will not print out verbose
information concerning per-process and per-device data. If you desire that
level of detail you can specify this option.
This option will output absolute block numbers
to three files prefixed by the specified output name:
prefix_device_r.dat
All read block numbers are output, first column is time (seconds), second is the
block number, and the third column is the ending block number.
prefix_device_w.dat
All write block numbers are output, first column is time (seconds), second is
the block number, and the third column is the ending block number.
prefix_device_c.dat
All block numbers (read and write) are output, first column is time (seconds),
second is the block number, and the third column is the ending block
number.
btt outputs a file containing Q and C
activity, the notion of active traces simply means that there are Q or C
traces occurring within a certain period of each other. The default values is
0.1 seconds; with this option allowing one to change that granularity. The
smaller the value, the more data points provided.
Normally, btt will produce data for all
devices detected in the traces parsed. With this option, one can reduce the
analysis to one or more devices provided in the string passed to this option.
The device identifiers are the major and minor number of each device, and each
device identifier is separated by a colon (:). A valid specifier for devices
8,0 and 8,8 would then be: 8,0:8,8.
The -e option supplies the list of executables
that will have I/Os analysed.
Shows a short summary of possible command line
option
Specifies the input file to analyse. This
should be a trace file produced by blktrace (8).
The -I option directs btt to output
iostat-like data to the specified file. Refer to the iostat (sysstat)
documentation for details on the data columns.
The -l option allows one to output per-IO D2C
latencies respectively. The supplied argument provides the basis for the
output name for each device.
The -L option allows one to output periodic
latency information for both Q2C and D2C latencies. The frequency specified
will regulate how often an average latency is output -- a floating point value
expressing seconds.
The -M option takes in a file generated by the
provided script (gen_disk_info.py), and allows for better output of device
names.
Specifies the output file name.
The -p option will generate a file that
contains a list of all IO "sequences" - showing the parts of each IO
(Q, A, I/M, D, & C).
The -P option will generate a file that
contains a list of all IO "sequences" - showing only the Q, D &
C operation times. The D & C time values are separated from the Q time
values with a vertical bar.
The -q option allows one to output per-IO Q2C
latencies respectively. The supplied argument provides the basis for the
output name for each device.
The -Q option allows one to output data files
showing the time stamp and the depth of active commands (those issued but not
completed).
Ignore remap traces; older kernels did not
implement the full remap PDU.
The -s option instructs btt to output seek
data, the argument provided is the basis for file names output. There are two
files per device, read seeks and write seeks.
The -S option specifies the interval to use
between data output, it defaults to once per second.
The -t/-T options allow one to set a start
and/or end time for analysing - analysing will only be done for traces after
-t's argument and before -T's argument. (-t and -T are optional, so if you
specify just -t, analysis will occur for all traces after the time specified.
Similarly, if only -T is specified, analysis stops after -T's seconds.)
This option instructs btt to generate a
data file containing histogram information for unplug traces on a per device
basis. It shows how many times an unplug was hit with a specified number of
IOs released. There are 21 output values into the file, as follows:
The file name(s) generated use the text string passed as an argument for the
prefix, followed by the device identifier in major,minor form, with a
.dat extension. For example, with -u up_hist specified on
the command line: up_hist_008,032.dat.
a value of 0 represents 0..4 counts
a value of 1 represents 5..9 counts
a value of 2 represents 10..14 counts
etc, until
a value of 20 represents 100+ counts
Shows the version of btt.
Requests a more verbose output.
The -z option allows one to output per-IO Q2D
latencies respectively. The supplied argument provides the basis for the
output name for each device.
AUTHORS¶
btt was written by Alan D. Brunelle. This man page was created from the blktrace documentation by Bas Zoetekouw.REPORTING BUGS¶
Report bugs to <linux-btrace@vger.kernel.org>COPYRIGHT¶
Copyright © 2006 Jens Axboe, Alan D. Brunelle and Nathan Scott.SEE ALSO¶
The btt Users Guide, which can be found in /usr/share/doc/blktrace/btt.pdf| September 29, 2007 | blktrace git-20070910192508 |