1 .TH "LTTNG" "1" "May 13th, 2014" "" ""
4 lttng \- LTTng 2.x tracer control command line tool
9 lttng [OPTIONS] <COMMAND>
13 The LTTng project aims at providing highly efficient tracing tools for Linux.
14 Its tracers help track down performance issues and debug problems
15 involving multiple concurrent processes and threads. Tracing across multiple
16 systems is also possible.
18 The \fBlttng\fP command line tool from the lttng-tools package is used to control
19 both kernel and user-space tracing. Every interaction with the tracer should
20 be done by this tool or by the liblttng-ctl library provided by the lttng-tools
23 LTTng uses a session daemon (lttng-sessiond(8)), acting as a tracing registry,
24 which allows you to interact with multiple tracers (kernel and user-space)
25 inside the same container, a tracing session. Traces can be gathered from the
26 kernel and/or instrumented applications (lttng-ust(3)). Aggregating and reading
27 those traces is done using the babeltrace(1) text viewer.
29 We introduce the notion of \fBtracing domains\fP which is essentially a type of
30 tracer (kernel, user space, JUL, LOG4J or Python for now). In the future, we
31 could see more tracer like for instance an hypervisor. For some commands,
32 you'll need to specify on which domain the command operates (\-u, \-k, \-l, \-j
33 or \-p). For instance, the kernel domain must be specified when enabling a
36 In order to trace the kernel, the session daemon needs to be running as root.
37 LTTng provides the use of a \fBtracing group\fP (default: tracing). Whomever is
38 in that group can interact with the root session daemon and thus trace the
39 kernel. Session daemons can co-exist, meaning that you can have a session daemon
40 running as Alice that can be used to trace her applications along side with a
41 root daemon or even a Bob daemon. We highly recommend starting the session
42 daemon at boot time for stable and long term tracing.
44 Each user-space application instrumented with lttng-ust(3) will automatically
45 register with the root session daemon and its user session daemon. This allows
46 each daemon to list the available traceable applications and tracepoints at any
47 given moment (See the \fBlist\fP command).
51 This program follows the usual GNU command line syntax with long options starting with
52 two dashes. Below is a summary of the available options.
57 Show summary of possible options and commands.
59 .BR "\-V, \-\-version"
62 .BR "\-v, \-\-verbose"
64 Three levels of verbosity are available which are triggered by putting additional v to
65 the option (\-vv or \-vvv)
68 Suppress all messages (even errors).
70 .BR "\-g, \-\-group NAME"
71 Set unix tracing group name. (default: tracing)
73 .BR "\-n, \-\-no-sessiond"
74 Don't automatically spawn a session daemon.
76 .BR "\-\-sessiond\-path PATH"
77 Set session daemon full binary path.
79 .BR "\-\-list\-options"
80 Simple listing of lttng options.
82 .BR "\-\-list\-commands"
83 Simple listing of lttng commands.
90 Machine interface (MI) mode converts the traditional pretty printing to a
91 machine output syntax. MI mode provides a format change-resistant way to access
92 information generated via the lttng command line.
94 When using MI mode, the data is printed on \fBstdout\fP. Error and warning are
95 printed on \fBstderr\fP with the pretty print default format.
97 If any errors occur during the execution of a command, the return value of the
98 command will be different than zero. In this case, lttng does NOT guarantee the
99 syntax and data validity of the generated MI output.
101 For XML output type, a schema definition (XSD) file used for validation can be
102 found under src/common/mi_lttng.xsd
107 \fBadd-context\fP [OPTIONS]
109 Add context to event(s) and/or channel(s).
111 A context is basically extra information appended to a channel. For instance,
112 you could ask the tracer to add the PID information for all events in a
113 channel. You can also add performance monitoring unit counters (perf PMU) using
116 For example, this command will add the context information 'prio' and two per-CPU
117 perf counters (hardware branch misses and cache misses), to all events in the trace
121 # lttng add-context \-k \-t prio \-t perf:cpu:branch-misses \\
122 \-t perf:cpu:cache-misses
125 Please take a look at the help (\-h/\-\-help) for a detailed list of available
128 Perf counters are available as per-CPU ("perf:cpu:...") and per-thread
129 ("perf:thread:...") counters. Currently, per-CPU counters can only be
130 used with the kernel tracing domain, and per-thread counters can only be
131 used with the UST tracing domain.
133 If no channel is given (\-c), the context is added to all channels that were
134 already enabled. If the session has no channel, a default channel is created.
135 Otherwise the context will be added only to the given channel (\-c).
137 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
144 Show summary of possible options and commands.
146 .BR "\-s, \-\-session NAME"
147 Apply on session name.
149 .BR "\-c, \-\-channel NAME"
150 Apply on channel name.
152 .BR "\-k, \-\-kernel"
153 Apply for the kernel tracer
155 .BR "\-u, \-\-userspace"
156 Apply for the user-space tracer
158 .BR "\-t, \-\-type TYPE"
159 Context type. You can repeat this option on the command line. Please
160 use "lttng add-context \-h" to list all available types.
165 \fBcalibrate\fP [OPTIONS]
167 Quantify LTTng overhead
169 The LTTng calibrate command can be used to find out the combined average
170 overhead of the LTTng tracer and the instrumentation mechanisms used. This
171 overhead can be calibrated in terms of time or using any of the PMU performance
172 counter available on the system.
174 For now, the only calibration implemented is that of the kernel function
175 instrumentation (kretprobes).
177 * Calibrate kernel function instrumentation
179 Let's use an example to show this calibration. We use an i7 processor with 4
180 general-purpose PMU registers. This information is available by issuing dmesg,
181 looking for "generic registers".
183 This sequence of commands will gather a trace executing a kretprobe hooked on
184 an empty function, gathering PMU counters LLC (Last Level Cache) misses
185 information (see lttng add-context \-\-help to see the list of available PMU
189 # lttng create calibrate-function
190 # lttng enable-event calibrate \-\-kernel \\
191 \-\-function lttng_calibrate_kretprobe
192 # lttng add-context \-\-kernel \-t perf:cpu:LLC-load-misses \\
193 \-t perf:cpu:LLC-store-misses \\
194 \-t perf:cpu:LLC-prefetch-misses
196 # for a in $(seq 1 10); do \\
197 lttng calibrate \-\-kernel \-\-function;
200 # babeltrace $(ls \-1drt ~/lttng-traces/calibrate-function-* \\
204 The output from babeltrace can be saved to a text file and opened in a
205 spreadsheet (e.g. oocalc) to focus on the per-PMU counter delta between
206 consecutive "calibrate_entry" and "calibrate_return" events. Note that these
207 counters are per-CPU, so scheduling events would need to be present to account
208 for migration between CPU. Therefore, for calibration purposes, only events
209 staying on the same CPU must be considered.
211 The average result, for the i7, on 10 samples:
215 perf_LLC_load_misses: 5.0 0.577
216 perf_LLC_store_misses: 1.6 0.516
217 perf_LLC_prefetch_misses: 9.0 14.742
220 As we can notice, the load and store misses are relatively stable across runs
221 (their standard deviation is relatively low) compared to the prefetch misses.
222 We can conclude from this information that LLC load and store misses can be
223 accounted for quite precisely, but prefetches within a function seems to behave
224 too erratically (not much causality link between the code executed and the CPU
225 prefetch activity) to be accounted for.
231 Show summary of possible options and commands.
233 .BR "\-k, \-\-kernel"
234 Apply for the kernel tracer
236 .BR "\-u, \-\-userspace"
237 Apply for the user-space tracer
240 Dynamic function entry/return probe (default)
245 \fBcreate\fP [NAME] [OPTIONS]
247 Create tracing session.
249 A tracing session contains channel(s) which contains event(s). It is domain
250 agnostic, meaning that channels and events can be enabled for the
251 user-space tracer and/or the kernel tracer. It acts like a container
252 aggregating multiple tracing sources.
254 On creation, a \fB.lttngrc\fP file is created in your $HOME directory
255 containing the current session name. If NAME is omitted, a session name is
256 automatically created having this form: 'auto-yyyymmdd-hhmmss'.
258 If no \fB\-o, \-\-output\fP is specified, the traces will be written in
261 The $HOME environment variable can be overridden by defining the environment
262 variable LTTNG_HOME. This is useful when the user running the commands has
263 a non-writeable home directory.
265 The session name MUST NOT contain the character '/'.
271 Show summary of possible options and commands.
273 .BR "\-\-list-options"
274 Simple listing of options
276 .BR "\-o, \-\-output PATH"
277 Specify output path for traces
280 Traces will not be output
283 Set the session in snapshot mode. Created in no-output mode and uses the
284 URL, if one is specified, as the default snapshot output. Every channel will be set
285 in overwrite mode and with mmap output (splice not supported).
287 .BR "\-\-live [USEC]"
288 Set the session exclusively in live mode. The parameter is the delay in micro
289 seconds before the data is flushed and streamed. The live mode allows you to
290 stream the trace and view it while it's being recorded by any tracer. For that,
291 you need a lttng-relayd and this session requires a network URL (\-U or
292 \-C/\-D). If no USEC nor URL is provided, the default is to use a timer value
293 set to 1000000 and the network URL set to net://127.0.0.1.
295 To read a live session, you can use babeltrace(1) or the live streaming
296 protocol in doc/live-reading-protocol.txt. Here is an example:
299 $ lttng-relayd -o /tmp/lttng
300 $ lttng create --live 200000 -U net://localhost
301 $ lttng enable-event -a --userspace
305 After the start, you'll be able to read the events while they are being
306 recorded in /tmp/lttng.
309 .BR "\-\-shm-path PATH"
311 Path where shared memory holding buffers should be created. Useful
312 when used with PRAMFS or other persistent memory filesystems to extract
313 trace data in the event of a crash requiring a reboot.
315 See the \fBlttng-crash(1)\fP utility for more information on crash recovery.
318 .BR "\-U, \-\-set-url=URL"
319 Set URL for the consumer output destination. It is persistent for the
320 session lifetime. Redo the command to change it. This will set both data
321 and control URL for network.
323 .BR "\-C, \-\-ctrl-url=URL"
324 Set control path URL. (Must use -D also)
326 .BR "\-D, \-\-data-url=URL"
327 Set data path URL. (Must use -C also)
329 Using these options, each API call can be controlled individually. For
330 instance, \-C does not enable the consumer automatically. You'll need the \-e
335 proto://[HOST|IP][:PORT1[:PORT2]][/TRACE_PATH]
337 Supported protocols are (proto):
340 Local filesystem full path.
344 This will use the default network transport layer which is TCP for both
345 control (PORT1) and data port (PORT2). The default ports are
346 respectively 5342 and 5343. Note that net[6]:// is not yet supported.
350 Can only be used with -C and -D together
352 NOTE: IPv6 address MUST be enclosed in brackets '[]' (rfc2732)
357 # lttng create -U net://192.168.1.42
359 Uses TCP and default ports for the given destination.
362 # lttng create -U net6://[fe80::f66d:4ff:fe53:d220]
364 Uses TCP, default ports and IPv6.
367 # lttng create s1 -U net://myhost.com:3229
369 Create session s1 and set its consumer to myhost.com on port 3229 for control.
374 \fBdestroy\fP [NAME] [OPTIONS]
376 Teardown tracing session
378 Free memory on the session daemon and tracer side. It's gone!
380 If NAME is omitted, the session name is taken from the .lttngrc file.
386 Show summary of possible options and commands.
391 .BR "\-\-list-options"
392 Simple listing of options
397 \fBenable-channel\fP NAME[,NAME2,...] (\-k | \-u) [OPTIONS]
399 Enable tracing channel
401 To enable an event, you must enable both the event and the channel that
404 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
407 Exactly one of \-k or -u must be specified.
409 It is important to note that if a certain type of buffers is used, the session
410 will be set with that type and all other subsequent channel needs to have the
413 Note that once the session has been started and enabled on the tracer side,
414 it's not possible anymore to enable a new channel for that session.
422 .BR "\-\-list-options"
423 Simple listing of options
425 .BR "\-s, \-\-session NAME"
426 Apply on session name
428 .BR "\-k, \-\-kernel"
429 Apply to the kernel tracer
431 .BR "\-u, \-\-userspace"
432 Apply to the user-space tracer
435 Discard event when subbuffers are full (default)
438 Flight recorder mode: overwrites events when subbuffers are full. The
439 number of subbuffer must be 2 or more.
441 .BR "\-\-subbuf-size SIZE"
442 Subbuffer size in bytes {+k,+M,+G}.
443 (default UST uid: 131072, UST pid: 4096, kernel: 262144, metadata: 4096)
444 Rounded up to the next power of 2.
446 The minimum subbuffer size, for each tracer, is the max value between
447 the default above and the system page size. You can issue this command
448 to get the current page size on your system: \fB$ getconf PAGE_SIZE\fP
450 .BR "\-\-num-subbuf NUM"
451 Number of subbuffers. (default UST uid: 4, UST pid: 4, kernel: 4,
452 metadata: 2) Rounded up to the next power of 2.
454 .BR "\-\-switch-timer USEC"
455 Switch subbuffer timer interval in µsec.
456 (default UST uid: 0, UST pid: 0, kernel: 0, metadata: 0)
458 .BR "\-\-read-timer USEC"
459 Read timer interval in µsec.
460 (default UST uid: 0, UST pid: 0, kernel: 200000, metadata: 0)
462 .BR "\-\-output TYPE"
463 Channel output type. Possible values: mmap, splice
464 (default UST uid: mmap, UST pid: mmap, kernel: splice, metadata: mmap)
466 .BR "\-\-buffers-uid"
467 Use per UID buffer (\-u only). Buffers are shared between applications
468 that have the same UID.
470 .BR "\-\-buffers-pid"
471 Use per PID buffer (\-u only). Each application has its own buffers.
473 .BR "\-\-buffers-global"
474 Use shared buffer for the whole system (\-k only)
476 .BR "\-C, \-\-tracefile-size SIZE"
477 Maximum size of each tracefile within a stream (in bytes).
478 0 means unlimited. (default: 0)
479 Note: traces generated with this option may inaccurately report
480 discarded events as of CTF 1.8.
482 .BR "\-W, \-\-tracefile-count COUNT"
483 Used in conjunction with \-C option, this will limit the number of files
484 created to the specified count. 0 means unlimited. (default: 0)
489 $ lttng enable-channel -k -C 4096 -W 32 chan1
491 For each stream, the maximum size of each trace file will be 4096 bytes and
492 there will be a maximum of 32 different files. The file count is appended after
493 the stream number as seen in the following example. The last trace file is
494 smaller than 4096 since it was not completely filled.
497 ~/lttng-traces/[...]/chan1_0_0 (4096)
498 ~/lttng-traces/[...]/chan1_0_1 (4096)
499 ~/lttng-traces/[...]/chan1_0_2 (3245)
500 ~/lttng-traces/[...]/chan1_1_0 (4096)
505 $ lttng enable-channel -k -C 4096
507 This will create trace files of 4096 bytes and will create new ones as long as
508 there is data available.
513 \fBenable-event\fP NAME[,NAME2,...] (\-k | \-u | \-j | \-l | \-p) [OPTIONS]
517 A tracing event is always assigned to a channel. If \fB\-c, \-\-channel\fP is
518 omitted, a default channel named '\fBchannel0\fP' is created and the event is
519 added to it. If \fB\-c, \-\-channel\fP is omitted, but a non-default
520 channel already exists within the session, an error is returned. For the
521 user-space tracer, using \fB\-a, \-\-all\fP is the same as using the
524 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
531 Show summary of possible options and commands.
533 .BR "\-\-list-options"
534 Simple listing of options
536 .BR "\-s, \-\-session NAME"
537 Apply on session name
539 .BR "\-c, \-\-channel NAME"
540 Apply on channel name
543 Enable all tracepoints and syscalls. This actually enables a single
546 .BR "\-k, \-\-kernel"
547 Apply for the kernel tracer
549 .BR "\-u, \-\-userspace"
550 Apply for the user-space tracer
553 Apply for Java application using Java Util Logging interface (JUL)
556 Apply for Java application using LOG4J
558 .BR "\-p, \-\-python"
559 Apply for Python application using the logging module.
562 Tracepoint event (default). Userspace tracer supports wildcards at the end
563 of string. Don't forget to quote to deal with bash expansion.
570 .BR "\-\-loglevel NAME"
571 Tracepoint loglevel range from 0 to loglevel. Listed in the help (\-h).
572 For the JUL domain, the loglevel ranges are detailed with the \-\-help
573 option thus starting from SEVERE to FINEST.
574 For the LOG4J domain, loglevels range from FATAL to TRACE which are also
575 detailed in the help.
576 For the Python domain, loglevels range from CRITICAL to DEBUG which are
577 detailed in the help as well.
579 .BR "\-\-loglevel-only NAME"
580 Tracepoint loglevel (only this loglevel).
581 The loglevel or loglevel-only options should be combined with a
582 tracepoint name or tracepoint wildcard.
584 .BR "\-\-probe (addr | symbol | symbol+offset)"
585 Dynamic probe. Addr and offset can be octal (0NNN...), decimal (NNN...)
586 or hexadecimal (0xNNN...)
588 .BR "\-\-function (addr | symbol | symbol+offset)"
589 Dynamic function entry/return probe. Addr and offset can be octal
590 (0NNN...), decimal (NNN...) or hexadecimal (0xNNN...)
595 .BR "\-\-filter 'expression'"
596 Set a filter on a newly enabled event. Filter expression on event
597 fields and context. The event will be recorded if the filter's
598 expression evaluates to TRUE. Only specify on first activation of a
599 given event within a session.
600 Specifying a filter is only allowed when enabling events within a session before
601 tracing is started. If the filter fails to link with the event
602 within the traced domain, the event will be discarded.
607 'intfield > 500 && intfield < 503'
608 '(strfield == "test" || intfield != 10) && intfield > 33'
609 'doublefield > 1.1 && intfield < 5.3'
613 Wildcards are allowed at the end of strings:
615 In string literals, the escape character is a '\\'. Use '\\*' for
616 the '*' character, and '\\\\' for the '\\' character sequence. Wildcard
617 matches any sequence of characters, including an empty sub-string
618 (matches 0 or more characters). Enumeration fields can currently only be
619 compared as integers.
621 Context information can be used for filtering. The examples below shows
622 usage of context filtering on the process name (using a wildcard), process ID
623 range, and unique thread ID. The process and thread IDs of
624 running applications can be found under columns "PID" and "LWP" of the
628 '$ctx.procname == "demo*"'
629 '$ctx.vpid >= 4433 && $ctx.vpid < 4455'
633 Context information is available to all filters whether or not the add-context
634 command has been used to add it to the event's channel, as long as the context
635 field exists for that domain. For example, the filter examples given above will
636 never fail to link: no add-context is required for the event's channel.
639 .BR "\-x, \-\-exclude LIST"
640 Add exclusions to UST tracepoints:
641 Events that match any of the items in the comma-separated LIST are not
642 enabled, even if they match a wildcard definition of the event.
644 This option is also applicable with the \fB\-a, \-\-all\fP option,
645 in which case all UST tracepoints are enabled except the ones whose
646 names match any of the items in LIST.
651 \fBdisable-channel\fP NAME[,NAME2,...] (\-k | \-u) [OPTIONS]
653 Disable tracing channel
655 Disabling a channel disables the tracing of all of the channel's events. A channel
656 can be re-enabled by calling \fBlttng enable-channel NAME\fP again.
658 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
665 Show summary of possible options and commands.
667 .BR "\-\-list-options"
668 Simple listing of options
670 .BR "\-s, \-\-session NAME"
671 Apply on session name
673 .BR "\-k, \-\-kernel"
674 Apply for the kernel tracer
676 .BR "\-u, \-\-userspace"
677 Apply for the user-space tracer
682 \fBdisable-event\fP NAME[,NAME2,...] (\-k | \-u | \-j | \-l | \-p) [TYPE] [OPTIONS]
684 Disable tracing event
686 The event, once disabled, can be re-enabled by calling \fBlttng enable-event
689 If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
692 If \fB\-c, \-\-channel\fP is omitted, the default channel name is used.
693 If \fB\-c, \-\-channel\fP is omitted, but a non-default channel already
694 exists within the session, an error is returned.
700 Show summary of possible options and commands.
702 .BR "\-\-list-options"
703 Simple listing of options
705 .BR "\-s, \-\-session NAME"
706 Apply on session name
708 .BR "\-c, \-\-channel NAME"
709 Apply on channel name
711 .BR "\-a, \-\-all-events"
712 Disable all events. This does NOT ONLY disable "*" but rather every known
713 events of the session
715 .BR "\-k, \-\-kernel"
716 Apply for the kernel tracer
718 .BR "\-u, \-\-userspace"
719 Apply for the user-space tracer
722 Apply for Java application using Java Util Logging interface (JUL)
725 Apply for Java application using LOG4J
727 .BR "\-p, \-\-python"
728 Apply for Python application using the logging module
731 .B TYPE (kernel domain only):
735 Disable event of all type
738 Disable event of type tracepoint
741 Disable event of type syscall
744 Disable event of type probe
747 Disable event of type function
752 \fBlist\fP [OPTIONS] [SESSION [SESSION OPTIONS]]
754 List tracing session information.
756 With no arguments, it will list available tracing session(s).
758 With the session name, it will display the details of the session including
759 the trace file path, the associated channels and their state (activated
760 and deactivated), the activated events and more.
762 With \-k alone, it will list all available kernel events (except the system
764 With \-j alone, the available JUL event from registered application will be
765 list. The event corresponds to the Logger name in the Java JUL application.
766 With \-l alone, the available LOG4J event from registered application will be
767 list. The event corresponds to the Logger name in the Java LOG4J application.
768 With \-p alone, the available Python event from registered application will be
769 list. The event corresponds to the Logger name in the Python application.
770 With \-u alone, it will list all available user-space events from registered
771 applications. Here is an example of 'lttng list \-u':
774 PID: 7448 - Name: /tmp/lttng-ust/tests/hello/.libs/lt-hello
775 ust_tests_hello:tptest_sighandler (type: tracepoint)
776 ust_tests_hello:tptest (type: tracepoint)
779 You can now enable any event listed by using the name :
780 \fBust_tests_hello:tptest\fP.
786 Show summary of possible options and commands.
788 .BR "\-\-list-options"
789 Simple listing of options
791 .BR "\-k, \-\-kernel"
794 .BR "\-u, \-\-userspace"
795 Select user-space domain.
798 Apply for Java application using JUL
801 Apply for Java application using LOG4J
803 .BR "\-p, \-\-python"
804 Apply for Python application using the logging module.
806 .BR "\-f, \-\-fields"
813 .BR "\-c, \-\-channel NAME"
814 List details of a channel
816 .BR "\-d, \-\-domain"
817 List available domain(s)
822 \fBload\fP [OPTIONS] [NAME]
824 Load tracing session configuration
826 If NAME is omitted, all session configurations found in both the user's session
827 configuration directory (default: ~/.lttng/sessions/) and the system session
828 configuration directory (default: /etc/lttng/sessions/) will be loaded. Note
829 that the sessions in the user directory are loaded first and then the system
830 wide directory are loaded.
836 Show summary of possible options and commands.
839 Load all session configurations (default).
841 .BR "\-i, \-\-input-path PATH"
842 Specify the input path for session configurations. This overrides the default
843 session configuration directory.
846 Overwrite current session configuration(s) if a session of the same name
852 \fBmetadata\fP [OPTIONS] ACTION
854 Metadata command for a LTTng session.
860 Show summary of possible options and commands.
862 .BR "\-\-list-options"
863 Simple listing of options
869 \fBregenerate\fP [-s <NAME>]
870 Regenerate the metadata of a session. This allows the user to regenerate the
871 metadata after a major NTP correction and that way update the clock offset from
872 epoch in the metadata. Only works on kernel, UST per-uid and non-live sessions.
877 \fBsave\fP [OPTIONS] [SESSION]
879 Save tracing session configuration
881 If SESSION is omitted, all session configurations will be saved to individual
882 \fB.lttng\fP files under the user's session configuration directory (default:
883 ~/.lttng/sessions/). The default session configuration file naming scheme is
886 For instance, a user in the tracing group saving a session from a root session
887 daemon will save it in her/his user directory.
893 Show summary of possible options and commands.
896 Save all session configurations (default).
898 .BR "\-o, \-\-output-path PATH"
899 Specify the output path for saved sessions. This overrides the default session
900 configuration directory.
903 Overwrite session configuration file if session name clashes.
908 \fBset-session\fP NAME [OPTIONS]
910 Set current session name
912 Will change the session name in the .lttngrc file.
918 Show summary of possible options and commands.
920 .BR "\-\-list-options"
921 Simple listing of options
926 \fBsnapshot\fP [OPTIONS] ACTION
928 Snapshot command for LTTng session.
934 Show summary of possible options and commands.
936 .BR "\-\-list-options"
937 Simple listing of options
943 \fBadd-output\fP [-m <SIZE>] [-s <NAME>] [-n <NAME>] <URL> | -C <URL> -D <URL>
945 Setup and add a snapshot output for a session. Output is the destination
946 where the snapshot will be sent. Only one output is permitted. To change it,
947 you'll need to delete it and add back the new one.
950 \fBdel-output\fP ID | NAME [-s <NAME>]
952 Delete an output for a session using the output's ID. You can either specify the
953 output by name or use its ID as returned by the list-output command.
956 \fBlist-output\fP [-s <NAME>]
958 List the output of a session. Attributes of the output are printed.
961 \fBrecord\fP [-m <SIZE>] [-s <NAME>] [-n <NAME>] [<URL> | -C <URL> -D <URL>]
963 Snapshot a session's buffer(s) for all domains. If an URL is specified, it is
964 used instead of a previously added output. Specifying only a name or/and a max
965 size will override the current output values. For instance, you can record a
966 snapshot with a custom maximum size or with a different name.
969 $ lttng snapshot add-output -n mysnapshot file:///data/snapshot
971 $ lttng snapshot record -n new_name_snapshot
974 The above will create a snapshot in /data/snapshot/new_name_snapshot* directory
975 rather then in mysnapshot*/
978 .B DETAILED ACTION OPTIONS
981 .BR "\-s, \-\-session NAME"
982 Apply to session name.
984 .BR "\-n, \-\-name NAME"
985 Name of the snapshot's output.
987 .BR "\-m, \-\-max-size SIZE"
988 Maximum size in bytes of the snapshot. The maximum size does not include the
989 metadata file. Human readable format is accepted: {+k,+M,+G}. For instance,
992 .BR "\-C, \-\-ctrl-url URL"
993 Set control path URL. (Must use -D also)
995 .BR "\-D, \-\-data-url URL"
996 Set data path URL. (Must use -C also)
1001 \fBstart\fP [NAME] [OPTIONS]
1005 It will start tracing for all tracers for a specific tracing session.
1006 If NAME is omitted, the session name is taken from the .lttngrc file.
1012 Show summary of possible options and commands.
1014 .BR "\-\-list-options"
1015 Simple listing of options
1020 \fBstop\fP [NAME] [OPTIONS]
1024 It will stop tracing for all tracers for a specific tracing session. Before
1025 returning, the command checks for data availability meaning that it will wait
1026 until the trace is readable for the session. Use \-\-no-wait to avoid this
1029 If NAME is omitted, the session name is taken from the .lttngrc file.
1035 Show summary of possible options and commands.
1037 .BR "\-\-list-options"
1038 Simple listing of options
1041 Don't wait for data availability.
1046 \fBtrack\fP (-k | -u) --pid [PID1[,PID2[,...]]] [OPTIONS]
1048 Adds one or more entries to a tracker
1050 The \fBtrack\fP command adds one or more entries to a tracker. A tracker is
1051 a whitelist of resources. Tracked resources are allowed to emit events, provided
1052 those events are enabled (see the \fBenable-event\fP command).
1054 Tracker entries can be removed from the whitelist with the
1055 \fBuntrack\fP command.
1057 As of this version, the only available tracker is the \fBPID tracker\fP. The
1058 process ID (PID) tracker follows one or more process IDs;
1059 only the processes with a tracked PID are allowed to emit events. By default,
1060 all possible PIDs on the system are tracked: any process may emit enabled
1061 events (equivalent of \fBlttng track \-\-pid \-\-all\fR for all domains).
1063 With the PID tracker, it is possible, for example, to record all system calls
1064 called by a given process:
1067 $ lttng enable-event --kernel --all --syscall
1068 $ lttng track --kernel --pid 2345
1072 If all the PIDs are tracked (i.e. \fBlttng track \-\-pid \-\-all\fR, which
1073 is the default state of all domains when creating a tracing session), then
1074 using the \fBtrack\fR command with one or more specific PIDs has the effect of
1075 first removing all the PIDs from the whitelist, then adding the specified PIDs.
1077 Assume the maximum PID is 7 for the following examples:
1080 Initial whitelist: [0] [1] [2] [3] [4] [5] [6] [7]
1082 $ lttng track --userspace --pid 3,6,7
1084 Whitelist: [ ] [ ] [ ] [3] [ ] [ ] [6] [7]
1086 $ lttng untrack --userspace --pid 7
1088 Whitelist: [ ] [ ] [ ] [3] [ ] [ ] [6] [ ]
1090 $ lttng track --userspace --pid 1,5
1092 Whitelist: [ ] [1] [ ] [3] [ ] [5] [6] [ ]
1095 It should be noted that the PID tracker tracks the numeric process IDs.
1096 Should a process with a given ID exit and another process be given this
1097 ID, then the latter would also be allowed to emit events.
1099 See the \fBuntrack\fR command's documentation for more details about
1105 .BR "\-s, \-\-session NAME"
1106 Apply to session name.
1108 .BR "\-k, \-\-kernel"
1109 Apply to the kernel tracer.
1111 .BR "\-u, \-\-userspace"
1112 Apply to the user space tracer.
1114 .BR "\-p, \-\-pid [PIDS]"
1115 Track process IDs PIDS (add to whitelist).
1117 PIDS is a comma-separated list of PIDs to add to the PID tracker.
1119 The PIDS argument must be omitted when also using the \fB\-\-all\fP option.
1122 Used in conjunction with an empty \fB\-\-pid\fP option: track all process IDs
1123 (add all entries to whitelist).
1126 Show summary of possible options and commands.
1128 .BR "\-\-list-options"
1129 Simple listing of options
1134 \fBuntrack\fP (-k | -u) --pid [PID1[,PID2[,...]]] [OPTIONS]
1136 Removes one or more entries from a tracker
1138 See the \fBtrack\fP command's documentation to learn more about LTTng
1141 The \fBuntrack\fP command removes specific resources from a tracker. The
1142 resources to remove must have been precedently added by the
1143 \fBtrack\fP command. It is also possible to remove all the resources
1144 from the whitelist using the \fB\-\-all\fR option.
1146 As of this version, the only available tracker is the \fBPID tracker\fP.
1148 One common operation is to create a tracing session, remove all the entries
1149 from the PID tracker whitelist, start tracing, and then manually track PIDs
1150 while tracing is active.
1152 Assume the maximum PID is 7 for the following examples:
1157 Initial whitelist: [0] [1] [2] [3] [4] [5] [6] [7]
1159 $ lttng untrack --userspace --pid --all
1161 Whitelist: [ ] [ ] [ ] [ ] [ ] [ ] [ ] [ ]
1163 $ lttng enable-event --userspace ...
1166 $ lttng track --userspace --pid 3,5
1168 Whitelist: [ ] [ ] [ ] [3] [ ] [5] [ ] [ ]
1170 $ lttng track --userspace --pid 2
1172 Whitelist: [ ] [ ] [2] [3] [ ] [5] [ ] [ ]
1175 See the \fBtrack\fR command's documentation for more details about
1181 .BR "\-s, \-\-session NAME"
1182 Apply to session name.
1184 .BR "\-k, \-\-kernel"
1185 Apply to the kernel tracer.
1187 .BR "\-u, \-\-userspace"
1188 Apply to the user space tracer.
1190 .BR "\-p, \-\-pid [PIDS]"
1191 Stop tracking process IDs PIDS (remove from whitelist).
1193 PIDS is a comma-separated list of PIDs to remove from the PID tracker.
1195 The PIDS argument must be omitted when also using the \fB\-\-all\fP option.
1198 Used in conjunction with an empty \fB\-\-pid\fP option: stop tracking all
1199 process IDs (remove all entries from whitelist).
1202 Show summary of possible options and commands.
1204 .BR "\-\-list-options"
1205 Simple listing of options
1212 Show version information
1218 Show summary of possible options and commands.
1220 .BR "\-\-list-options"
1221 Simple listing of options
1226 \fBview\fP [SESSION_NAME] [OPTIONS]
1228 View traces of a tracing session. By default, the babeltrace viewer
1229 will be used for text viewing. If SESSION_NAME is omitted, the session
1230 name is taken from the .lttngrc file.
1238 .BR "\-\-list-options"
1239 Simple listing of options
1241 .BR "\-t, \-\-trace-path PATH"
1242 Trace directory path for the viewer
1244 .BR "\-e, \-\-viewer CMD"
1245 Specify viewer and/or options to use This will completely override the
1246 default viewers so please make sure to specify the full command. The
1247 trace directory path of the session will be appended at the end to the
1252 .SH "JUL/LOG4J DOMAIN"
1254 This section explains the JUL and LOG4J domain where JUL stands for Java Util
1255 Logging. You can use these by using the \fBliblttng-ust-<domain>-jni.so\fP from
1256 the lttng-ust(3) project.
1258 The LTTng Java Agent uses JNI to link the UST tracer to the Java application
1259 that uses the agent. Thus, it behaves similarly to the UST domain (\-u). When
1260 enabling events, you enable a Logger name that will then be mapped to a default
1261 UST tracepoint called \fBlttng_jul:<domain>_event\fP in the
1262 \fBlttng_<domain>_channel\fP. Using the lttng-ctl API, any JUL/LOG4J events
1263 must use the tracepoint event type (same as \-\-tracepoint).
1265 Because of the default immutable channel, the \fBenable-channel\fP command CAN
1266 NOT be used with the JUL and LOG4J domain thus not having any options.
1268 Also, loglevels are supported. Use \fBlttng enable-event \-h\fP to list them.
1269 Wildcards are NOT supported except the "*" meaning all events (same as \-a).
1271 Exactly like the UST domain, if the Java application has the same UID as you,
1272 you can trace it. Same goes for the tracing group accessing root applications.
1274 Finally, you can list every Logger name that are available from registered
1275 applications to the session daemon by using \fBlttng list \-j\fP or \fB\-l\fP.
1277 Here is an example on how to use the JUL domain.
1282 $ lttng create aSession
1283 $ lttng enable-event -s aSession -j MyCustomLoggerName
1287 More information can be found in the lttng-ust documentation, see
1288 java-util-logging.txt
1293 On success 0 is returned and a positive value on error. Value of 1 means a command
1294 error, 2 an undefined command, 3 a fatal error and 4 a command warning meaning that
1295 something went wrong during the command.
1297 Any other value above 10, please refer to
1298 .BR "<lttng/lttng-error.h>"
1299 for a detailed list or use lttng_strerror() to get a human readable string of
1303 .SH "ENVIRONMENT VARIABLES"
1306 Note that all command line options override environment variables.
1310 .IP "LTTNG_SESSIOND_PATH"
1311 Allows one to specify the full session daemon binary path to lttng command line
1312 tool. You can also use \-\-sessiond-path option having the same effect.
1316 .IP "LTTNG_SESSION_CONFIG_XSD_PATH"
1317 Set the path in which the \fBsession.xsd\fP session configuration schema may be
1324 .BR lttng-sessiond(8),
1325 .BR lttng-relayd(8),
1331 If you encounter any issues or usability problem, please report it on our
1332 mailing list <lttng-dev@lists.lttng.org> to help improve this project or
1333 at https://bugs.lttng.org which is a bug tracker.
1339 lttng is distributed under the GNU General Public License version 2. See the file
1340 COPYING for details.
1342 A Web site is available at http://lttng.org for more information on the LTTng
1345 You can also find our git tree at http://git.lttng.org.
1347 Mailing lists for support and development: <lttng-dev@lists.lttng.org>.
1349 You can find us on IRC server irc.oftc.net (OFTC) in #lttng.
1354 Thanks to Yannick Brosseau without whom this project would never have been so
1355 lean and mean! Also thanks to the Ericsson teams working on tracing which
1356 helped us greatly with detailed bug reports and unusual test cases.
1358 Thanks to our beloved packager Alexandre Montplaisir-Goncalves (Ubuntu and PPA
1359 maintainer) and Jon Bernard for our Debian packages.
1361 Special thanks to Michel Dagenais and the DORSAL laboratory at Polytechnique de
1362 Montreal for the LTTng journey.
1367 lttng-tools was originally written by Mathieu Desnoyers, Julien Desfossez and
1368 David Goulet. More people have since contributed to it. It is currently
1369 maintained by Jérémie Galarneau <jeremie.galarneau@efficios.com>.