1 .TH "LTTNG" "1" "December 3rd, 2012" "" ""
4 lttng \(em LTTng 2.1.x tracer control command line tool
10 lttng [OPTIONS] <COMMAND>
15 The LTTng project aims at providing highly efficient tracing tools for Linux.
16 It's tracers help tracking down performance issues and debugging problems
17 involving multiple concurrent processes and threads. Tracing across multiple
18 systems is also possible.
20 The \fBlttng\fP command line tool from the lttng-tools package is used to control
21 both kernel and user-space tracing. Every interactions with the tracer should
22 be done by this tool or by the liblttng-ctl provided with the lttng-tools
25 LTTng uses a session daemon (lttng-sessiond(8)), acting as a tracing registry,
26 which allows you to interact with multiple tracers (kernel and user-space)
27 inside the same container, a tracing session. Traces can be gathered from the
28 kernel and/or instrumented applications (lttng-ust(3)). Aggregating and reading
29 those traces is done using the babeltrace(1) text viewer.
31 We introduce the notion of \fBtracing domains\fP which is essentially a type of
32 tracer (kernel or user space for now). In the future, we could see a third
33 tracer being for instance an hypervisor. For some commands, you'll need to
34 specify on which domain the command applies (-u or -k). For instance, enabling
35 a kernel event, you must specify the kernel domain to the command so we know
36 for which tracer this event is for.
38 In order to trace the kernel, the session daemon needs to be running as root.
39 LTTng provides the use of a \fBtracing group\fP (default: tracing). Whomever is
40 in that group can interact with the root session daemon and thus trace the
41 kernel. Session daemons can co-exist meaning that you can have a session daemon
42 running as Alice that can be used to trace her applications along side with a
43 root daemon or even a Bob daemon. We highly recommend to start the session
44 daemon at boot time for stable and long term tracing.
46 Every user-space applications instrumented with lttng-ust(3), will
47 automatically register to the session daemon. This feature gives you the
48 ability to list available traceable applications and tracepoints on a per user
49 basis. (See \fBlist\fP command).
53 This program follow the usual GNU command line syntax with long options starting with
54 two dashes. Below is a summary of the available options.
59 Show summary of possible options and commands.
61 .BR "\-v, \-\-verbose"
63 Three levels of verbosity are available which are triggered by putting additional v to
64 the option (\-vv or \-vvv)
67 Suppress all messages (even errors).
69 .BR "\-g, \-\-group NAME"
70 Set unix tracing group name. (default: tracing)
72 .BR "\-n, \-\-no-sessiond"
73 Don't automatically spawn a session daemon.
75 .BR "\-\-sessiond\-path PATH"
76 Set session daemon full binary path.
78 .BR "\-\-list\-options"
79 Simple listing of lttng options.
81 .BR "\-\-list\-commands"
82 Simple listing of lttng commands.
88 Add context to event(s) and/or channel(s).
90 A context is basically extra information appended to a channel. For instance,
91 you could ask the tracer to add the PID information for all events in a
92 channel. You can also add performance monitoring unit counters (perf PMU) using
95 For example, this command will add the context information 'prio' and two perf
96 counters (hardware branch misses and cache misses), to all events in the trace
99 # lttng add-context \-k \-t prio \-t perf:branch-misses \-t perf:cache-misses
101 Please take a look at the help (\-h/\-\-help) for a detailed list of available
104 If no channel is given (\-c), the context is added to all channels. Otherwise
105 the context will be added only to the given channel (\-c).
107 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
115 Show summary of possible options and commands.
116 \-s, \-\-session NAME
117 Apply on session name.
118 \-c, \-\-channel NAME
119 Apply on channel name.
121 Apply for the kernel tracer
123 Apply for the user-space tracer
125 Context type. You can repeat this option on the command line. Please
126 use "lttng add-context \-h" to list all available types.
131 .IP "\fBcalibrate\fP"
133 Quantify LTTng overhead
135 The LTTng calibrate command can be used to find out the combined average
136 overhead of the LTTng tracer and the instrumentation mechanisms used. This
137 overhead can be calibrated in terms of time or using any of the PMU performance
138 counter available on the system.
140 For now, the only calibration implemented is that of the kernel function
141 instrumentation (kretprobes).
143 * Calibrate kernel function instrumentation
145 Let's use an example to show this calibration. We use an i7 processor with 4
146 general-purpose PMU registers. This information is available by issuing dmesg,
147 looking for "generic registers".
149 This sequence of commands will gather a trace executing a kretprobe hooked on
150 an empty function, gathering PMU counters LLC (Last Level Cache) misses
151 information (see lttng add-context \-\-help to see the list of available PMU
154 # lttng create calibrate-function
155 # lttng enable-event calibrate \-\-kernel \-\-function lttng_calibrate_kretprobe
156 # lttng add-context \-\-kernel \-t perf:LLC-load-misses \-t perf:LLC-store-misses \\
157 \-t perf:LLC-prefetch-misses
159 # for a in $(seq 1 10); do \\
160 lttng calibrate \-\-kernel \-\-function;
163 # babeltrace $(ls \-1drt ~/lttng-traces/calibrate-function-* | tail \-n 1)
165 The output from babeltrace can be saved to a text file and opened in a
166 spreadsheet (e.g. oocalc) to focus on the per-PMU counter delta between
167 consecutive "calibrate_entry" and "calibrate_return" events. Note that these
168 counters are per-CPU, so scheduling events would need to be present to account
169 for migration between CPU. Therefore, for calibration purposes, only events
170 staying on the same CPU must be considered.
172 The average result, for the i7, on 10 samples:
175 perf_LLC_load_misses: 5.0 0.577
176 perf_LLC_store_misses: 1.6 0.516
177 perf_LLC_prefetch_misses: 9.0 14.742
179 As we can notice, the load and store misses are relatively stable across runs
180 (their standard deviation is relatively low) compared to the prefetch misses.
181 We can conclude from this information that LLC load and store misses can be
182 accounted for quite precisely, but prefetches within a function seems to behave
183 too erratically (not much causality link between the code executed and the CPU
184 prefetch activity) to be accounted for.
191 Show summary of possible options and commands.
193 Apply for the kernel tracer
195 Apply for the user-space tracer
197 Dynamic function entry/return probe (default)
202 .IP "\fBcreate\fP [NAME] [OPTIONS]
204 Create tracing session.
206 A tracing session contains channel(s) which contains event(s). It is domain
207 agnostic meaning that you can enable channels and events for either the
208 user-space tracer and/or the kernel tracer. It acts like a container
209 aggregating multiple tracing sources.
211 On creation, a \fB.lttngrc\fP file is created in your $HOME directory
212 containing the current session name. If NAME is omitted, a session name is
213 automatically created having this form: 'auto-yyyymmdd-hhmmss'.
215 If no \fB\-o, \-\-output\fP is specified, the traces will be written in
223 Show summary of possible options and commands.
225 Simple listing of options
227 Specify output path for traces
229 Using these options, each API call can be controlled individually. For
230 instance, \-C does not enable the consumer automatically. You'll need the \-e
234 Set URL for the consumer output destination. It is persistent for the
235 session lifetime. Redo the command to change it. This will set both
236 data and control URL for network.
237 \-C, \-\-ctrl-url=URL
238 Set control path URL. (Must use -D also)
239 \-D, \-\-data-url=URL
240 Set data path URL. (Must use -C also)
244 proto://[HOST|IP][:PORT1[:PORT2]][/TRACE_PATH]
246 Supported protocols are (proto):
248 Local filesystem full path.
251 This will use the default network transport layer which is TCP for both
252 control (PORT1) and data port (PORT2). The default ports are
253 respectively 5342 and 5343. Note that net[6]:// is not yet supported.
256 Can only be used with -C and -D together
258 NOTE: IPv6 address MUST be enclosed in brackets '[]' (rfc2732)
262 # lttng create -U net://192.168.1.42
263 Uses TCP and default ports for the given destination.
265 # lttng create -U net6://[fe80::f66d:4ff:fe53:d220]
266 Uses TCP, default ports and IPv6.
268 # lttng create s1 -U net://myhost.com:3229
269 Create session s1 and set its consumer to myhost.com on port 3229 for control.
274 .IP "\fBdestroy\fP [OPTIONS] [NAME]"
276 Teardown tracing session
278 Free memory on the session daemon and tracer side. It's gone!
280 If NAME is omitted, the session name is taken from the .lttngrc file.
287 Show summary of possible options and commands.
291 Simple listing of options
296 .IP "\fBenable-channel\fP NAME[,NAME2,...] [-k|-u] [OPTIONS]"
298 Enable tracing channel
300 To enable an event, you must enable both the event and the channel that
303 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
306 It is important to note that if a certain type of buffers is used, the session
307 will be set with that type and all other subsequent channel need to have the
317 Simple listing of options
318 \-s, \-\-session NAME
319 Apply on session name
321 Apply to the kernel tracer
323 Apply to the user-space tracer
326 Discard event when subbuffers are full (default)
328 Flight recorder mode : overwrites events when subbuffers are full
330 Subbuffer size in bytes (default: 4096, kernel default: 262144)
331 Needs to be a power of 2 for both tracers
333 Number of subbuffers (default: 4)
334 Needs to be a power of 2 for both tracers
335 \-\-switch-timer USEC
336 Switch subbuffer timer interval in µsec (default: 0)
338 Read timer interval in µsec (default: 200)
340 Channel output type. Possible values: mmap, splice
342 Use per UID buffer (\-u only). Buffers are shared between applications
343 that have the same UID.
345 Use per PID buffer (\-u only). Each application has its own buffers.
347 Use shared buffer for the whole system (\-k only)
352 .IP "\fBenable-event\fP NAME[,NAME2,...] [-k|-u] [OPTIONS]"
356 A tracing event is always assigned to a channel. If \fB\-c, \-\-channel\fP is
357 omitted, a default channel named '\fBchannel0\fP' is created and the event is
358 added to it. For the user-space tracer, using \fB\-a, \-\-all\fP is the same as
359 using the wildcard "*".
361 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
369 Show summary of possible options and commands.
371 Simple listing of options
372 \-s, \-\-session NAME
373 Apply on session name
374 \-c, \-\-channel NAME
375 Apply on channel name
377 Enable all tracepoints and syscalls. This actually enable a single
380 Apply for the kernel tracer
382 Apply for the user-space tracer
385 Tracepoint event (default)
386 - userspace tracer supports wildcards at end of string. Don't forget to
387 quote to deal with bash expansion.
392 Tracepoint loglevel range from 0 to loglevel. Listed in the help (\-h).
393 \-\-loglevel-only NAME
394 Tracepoint loglevel (only this loglevel).
396 The loglevel or loglevel-only options should be combined with a
397 tracepoint name or tracepoint wildcard.
398 \-\-probe [addr | symbol | symbol+offset]
399 Dynamic probe. Addr and offset can be octal (0NNN...), decimal (NNN...)
400 or hexadecimal (0xNNN...)
401 \-\-function [addr | symbol | symbol+offset]
402 Dynamic function entry/return probe. Addr and offset can be octal
403 (0NNN...), decimal (NNN...) or hexadecimal (0xNNN...)
405 System call event. Enabling syscalls tracing (kernel tracer), you will
406 not be able to disable them with disable-event. This is a known
407 limitation. You can disable the entire channel to do the trick.
409 \-\-filter 'expression'
410 Set a filter on a newly enabled event. Filter expression on event
411 fields and context. Event recording depends on evaluation. Only
412 specify on first activation of a given event within a session.
413 Filter only allowed when enabling events within a session before
414 tracing is started. If the filter fails to link with the event
415 within the traced domain, the event will be discarded.
416 Currently, filter is only implemented for the user-space tracer.
420 'intfield > 500 && intfield < 503'
421 '(stringfield == "test" || intfield != 10) && intfield > 33'
422 'doublefield > 1.1 && intfield < 5.3'
424 Wildcards are allowed at the end of strings:
426 In string literals, the escape character is a '\\'. Use '\\*' for
427 the '*' character, and '\\\\' for the '\\' character. Wildcard
428 match any sequence of characters, including an empty sub-string
429 (match 0 or more characters).
431 Context information can be used for filtering. The examples
432 below show usage of context filtering on process name (with a
433 wildcard), process ID range, and unique thread ID for filtering.
434 The process and thread ID of running applications can be found
435 under columns "PID" and "LWP" of the "ps -eLf" command.
437 '$ctx.procname == "demo*"'
438 '$ctx.vpid >= 4433 && $ctx.vpid < 4455'
442 .IP "\fBdisable-channel\fP NAME[,NAME2,...] [\-k|\-u] [OPTIONS]"
444 Disable tracing channel
446 Disabling a channel makes all event(s) in that channel to stop tracing. You can
447 enable it back by calling \fBlttng enable-channel NAME\fP again.
449 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
457 Show summary of possible options and commands.
459 Simple listing of options
460 \-s, \-\-session NAME
461 Apply on session name
463 Apply for the kernel tracer
465 Apply for the user-space tracer
468 .IP "\fBdisable-event\fP NAME[,NAME2,...] [\-k|\-u] [OPTIONS]"
470 Disable tracing event
472 The event, once disabled, can be re-enabled by calling \fBlttng enable-event
475 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
483 Show summary of possible options and commands.
485 Simple listing of options
486 \-s, \-\-session NAME
487 Apply on session name
489 Disable all events. This does NOT disable "*" but rather
490 every known events of the session.
492 Apply for the kernel tracer
494 Apply for the user-space tracer
497 .IP "\fBlist\fP [\-k|\-u] [SESSION [SESSION_OPTIONS]]"
499 List tracing session information.
501 With no arguments, it will list available tracing session(s).
503 With the session name, it will display the details of the session including
504 the trace file path, the associated channels and their state (activated
505 and deactivated), the activated events and more.
507 With \-k alone, it will list all available kernel events (except the system
509 With \-u alone, it will list all available user-space events from registered
510 applications. Here is an example of 'lttng list \-u':
512 PID: 7448 - Name: /tmp/lttng-ust/tests/hello/.libs/lt-hello
513 ust_tests_hello:tptest_sighandler (type: tracepoint)
514 ust_tests_hello:tptest (type: tracepoint)
516 You can now enable any event listed by using the name :
517 \fBust_tests_hello:tptest\fP.
524 Show summary of possible options and commands.
526 Simple listing of options
530 Select user-space domain.
534 \-c, \-\-channel NAME
535 List details of a channel
537 List available domain(s)
540 .IP "\fBset-session\fP NAME"
542 Set current session name
544 Will change the session name in the .lttngrc file.
551 Show summary of possible options and commands.
553 Simple listing of options
558 .IP "\fBstart\fP [NAME] [OPTIONS]"
562 It will start tracing for all tracers for a specific tracing session.
564 If NAME is omitted, the session name is taken from the .lttngrc file.
571 Show summary of possible options and commands.
573 Simple listing of options
578 .IP "\fBstop\fP [NAME] [OPTIONS]"
582 It will stop tracing for all tracers for a specific tracing session. Before
583 returning, the command checks for data availability meaning that it will wait
584 until the trace is readable for the session. Use \-\-no-wait to avoid this
587 If NAME is omitted, the session name is taken from the .lttngrc file.
594 Show summary of possible options and commands.
596 Simple listing of options
598 Don't wait for data availability.
605 Show version information
612 Show summary of possible options and commands.
614 Simple listing of options
619 .IP "\fBview\fP [SESSION_NAME] [OPTIONS]"
621 View traces of a tracing session
623 By default, the babeltrace viewer will be used for text viewing.
625 If SESSION_NAME is omitted, the session name is taken from the .lttngrc file.
635 Simple listing of options
636 \-t, \-\-trace-path PATH
637 Trace directory path for the viewer
639 Specify viewer and/or options to use
640 This will completely override the default viewers so
641 please make sure to specify the full command. The trace
642 directory path of the session will be appended at the end
647 On success 0 is returned and a positive value on error. Value of 1 means a command
648 error, 2 an undefined command, 3 a fatal error and 4 a command warning meaning that
649 something went wrong during the command.
651 Any other value above 10, please refer to
652 .BR <lttng/lttng-error.h>
653 for a detailed list or use lttng_strerror() to get a human readable string of
657 .SH "ENVIRONMENT VARIABLES"
660 Note that all command line options override environment variables.
664 .IP "LTTNG_SESSIOND_PATH"
665 Allows one to specify the full session daemon binary path to lttng command line
666 tool. You can also use \-\-sessiond-path option having the same effect.
670 .BR lttng-sessiond(8),
672 .BR lttng-health-check(3)
675 If you encounter any issues or usability problem, please report it on our
676 mailing list <lttng-dev@lists.lttng.org> to help improve this project or
677 at https://bugs.lttng.org which is a bugtracker.
681 lttng is distributed under the GNU General Public License version 2. See the file
684 A Web site is available at http://lttng.org for more information on the LTTng
687 You can also find our git tree at http://git.lttng.org.
689 Mailing lists for support and development: <lttng-dev@lists.lttng.org>.
691 You can find us on IRC server irc.oftc.net (OFTC) in #lttng.
696 Thanks to Yannick Brosseau without whom this project would never have been so
697 lean and mean! Also thanks to the Ericsson teams working on tracing which
698 helped us greatly with detailed bug reports and unusual test cases.
700 Thanks to our beloved packager Alexandre Montplaisir-Goncalves (Ubuntu and PPA
701 maintainer) and Jon Bernard for our Debian packages.
703 Special thanks to Michel Dagenais and the DORSAL laboratory at Polytechnique de
704 Montreal for the LTTng journey.
709 lttng-tools was originally written by Mathieu Desnoyers, Julien Desfossez and
710 David Goulet. More people have since contributed to it. It is currently
711 maintained by David Goulet <dgoulet@efficios.com>.