Fix: grab more than one packet for snapshots
[lttng-tools.git] / doc / man / lttng.1
... / ...
CommitLineData
1.TH "LTTNG" "1" "May 13th, 2014" "" ""
2
3.SH "NAME"
4lttng \(em LTTng 2.x tracer control command line tool
5
6.SH "SYNOPSIS"
7
8.PP
9lttng [OPTIONS] <COMMAND>
10.SH "DESCRIPTION"
11
12.PP
13The LTTng project aims at providing highly efficient tracing tools for Linux.
14Its tracers help track down performance issues and debug problems
15involving multiple concurrent processes and threads. Tracing across multiple
16systems is also possible.
17
18The \fBlttng\fP command line tool from the lttng-tools package is used to control
19both kernel and user-space tracing. Every interaction with the tracer should
20be done by this tool or by the liblttng-ctl library provided by the lttng-tools
21package.
22
23LTTng uses a session daemon (lttng-sessiond(8)), acting as a tracing registry,
24which allows you to interact with multiple tracers (kernel and user-space)
25inside the same container, a tracing session. Traces can be gathered from the
26kernel and/or instrumented applications (lttng-ust(3)). Aggregating and reading
27those traces is done using the babeltrace(1) text viewer.
28
29We introduce the notion of \fBtracing domains\fP which is essentially a type of
30tracer (kernel, user space, JUL, LOG4J or Python for now). In the future, we
31could see more tracer like for instance an hypervisor. For some commands,
32you'll need to specify on which domain the command operates (\-u, \-k, \-l, \-j
33or \-p). For instance, the kernel domain must be specified when enabling a
34kernel event.
35
36In order to trace the kernel, the session daemon needs to be running as root.
37LTTng provides the use of a \fBtracing group\fP (default: tracing). Whomever is
38in that group can interact with the root session daemon and thus trace the
39kernel. Session daemons can co-exist, meaning that you can have a session daemon
40running as Alice that can be used to trace her applications along side with a
41root daemon or even a Bob daemon. We highly recommend starting the session
42daemon at boot time for stable and long term tracing.
43
44Each user-space application instrumented with lttng-ust(3) will automatically
45register with the root session daemon and its user session daemon. This allows
46each daemon to list the available traceable applications and tracepoints at any
47given moment (See the \fBlist\fP command).
48.SH "OPTIONS"
49
50.PP
51This program follow the usual GNU command line syntax with long options starting with
52two dashes. Below is a summary of the available options.
53.PP
54
55.TP
56.BR "\-h, \-\-help"
57Show summary of possible options and commands.
58.TP
59.BR "\-v, \-\-verbose"
60Increase verbosity.
61Three levels of verbosity are available which are triggered by putting additional v to
62the option (\-vv or \-vvv)
63.TP
64.BR "\-q, \-\-quiet"
65Suppress all messages (even errors).
66.TP
67.BR "\-g, \-\-group NAME"
68Set unix tracing group name. (default: tracing)
69.TP
70.BR "\-n, \-\-no-sessiond"
71Don't automatically spawn a session daemon.
72.TP
73.BR "\-\-sessiond\-path PATH"
74Set session daemon full binary path.
75.TP
76.BR "\-\-list\-options"
77Simple listing of lttng options.
78.TP
79.BR "\-\-list\-commands"
80Simple listing of lttng commands.
81.TP
82.BR "\-m, \-\-mi TYPE
83Machine interface
84
85TYPE supported: XML
86
87Machine interface (MI) mode converts the traditional pretty printing to a
88machine output syntax. MI mode provides a format change-resistant way to access
89information generated via the lttng command line.
90
91When using MI mode, the data is printed on \fBstdout\fP. Error and warning are
92printed on \fBstderr\fP with the pretty print default format.
93
94If any errors occur during the execution of a command, the return value of the
95command will be different than zero. In this case, lttng does NOT guarantee the
96syntax and data validity of the generated MI output.
97
98For XML output type, a schema definition (XSD) file used for validation can be
99found under src/common/mi_lttng.xsd
100
101.SH "COMMANDS"
102
103.PP
104\fBadd-context\fP [OPTIONS]
105.RS
106Add context to event(s) and/or channel(s).
107
108A context is basically extra information appended to a channel. For instance,
109you could ask the tracer to add the PID information for all events in a
110channel. You can also add performance monitoring unit counters (perf PMU) using
111the perf kernel API.
112
113For example, this command will add the context information 'prio' and two per-CPU
114perf counters (hardware branch misses and cache misses), to all events in the trace
115data output:
116
117.nf
118# lttng add-context \-k \-t prio \-t perf:cpu:branch-misses \\
119 \-t perf:cpu:cache-misses
120.fi
121
122Please take a look at the help (\-h/\-\-help) for a detailed list of available
123contexts.
124
125Perf counters are available as per-CPU ("perf:cpu:...") and per-thread
126("perf:thread:...") counters. Currently, per-CPU counters can only be
127used with the kernel tracing domain, and per-thread counters can only be
128used with the UST tracing domain.
129
130If no channel is given (\-c), the context is added to all channels that were
131already enabled. If the session has no channel, a default channel is created.
132Otherwise the context will be added only to the given channel (\-c).
133
134If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
135file.
136
137.B OPTIONS:
138
139.TP
140.BR "\-h, \-\-help"
141Show summary of possible options and commands.
142.TP
143.BR "\-s, \-\-session NAME"
144Apply on session name.
145.TP
146.BR "\-c, \-\-channel NAME"
147Apply on channel name.
148.TP
149.BR "\-k, \-\-kernel"
150Apply for the kernel tracer
151.TP
152.BR "\-u, \-\-userspace"
153Apply for the user-space tracer
154.TP
155.BR "\-t, \-\-type TYPE"
156Context type. You can repeat this option on the command line. Please
157use "lttng add-context \-h" to list all available types.
158.RE
159.PP
160
161.PP
162\fBcalibrate\fP [OPTIONS]
163.RS
164Quantify LTTng overhead
165
166The LTTng calibrate command can be used to find out the combined average
167overhead of the LTTng tracer and the instrumentation mechanisms used. This
168overhead can be calibrated in terms of time or using any of the PMU performance
169counter available on the system.
170
171For now, the only calibration implemented is that of the kernel function
172instrumentation (kretprobes).
173
174* Calibrate kernel function instrumentation
175
176Let's use an example to show this calibration. We use an i7 processor with 4
177general-purpose PMU registers. This information is available by issuing dmesg,
178looking for "generic registers".
179
180This sequence of commands will gather a trace executing a kretprobe hooked on
181an empty function, gathering PMU counters LLC (Last Level Cache) misses
182information (see lttng add-context \-\-help to see the list of available PMU
183counters).
184
185.nf
186# lttng create calibrate-function
187# lttng enable-event calibrate \-\-kernel \\
188 \-\-function lttng_calibrate_kretprobe
189# lttng add-context \-\-kernel \-t perf:cpu:LLC-load-misses \\
190 \-t perf:cpu:LLC-store-misses \\
191 \-t perf:cpu:LLC-prefetch-misses
192# lttng start
193# for a in $(seq 1 10); do \\
194 lttng calibrate \-\-kernel \-\-function;
195 done
196# lttng destroy
197# babeltrace $(ls \-1drt ~/lttng-traces/calibrate-function-* \\
198 | tail \-n 1)
199.fi
200
201The output from babeltrace can be saved to a text file and opened in a
202spreadsheet (e.g. oocalc) to focus on the per-PMU counter delta between
203consecutive "calibrate_entry" and "calibrate_return" events. Note that these
204counters are per-CPU, so scheduling events would need to be present to account
205for migration between CPU. Therefore, for calibration purposes, only events
206staying on the same CPU must be considered.
207
208The average result, for the i7, on 10 samples:
209
210.nf
211 Average Std.Dev.
212perf_LLC_load_misses: 5.0 0.577
213perf_LLC_store_misses: 1.6 0.516
214perf_LLC_prefetch_misses: 9.0 14.742
215.fi
216
217As we can notice, the load and store misses are relatively stable across runs
218(their standard deviation is relatively low) compared to the prefetch misses.
219We can conclude from this information that LLC load and store misses can be
220accounted for quite precisely, but prefetches within a function seems to behave
221too erratically (not much causality link between the code executed and the CPU
222prefetch activity) to be accounted for.
223
224.B OPTIONS:
225
226.TP
227.BR "\-h, \-\-help"
228Show summary of possible options and commands.
229.TP
230.BR "\-k, \-\-kernel"
231Apply for the kernel tracer
232.TP
233.BR "\-u, \-\-userspace"
234Apply for the user-space tracer
235.TP
236.BR "\-\-function"
237Dynamic function entry/return probe (default)
238.RE
239.PP
240
241.PP
242\fBcreate\fP [NAME] [OPTIONS]
243.RS
244Create tracing session.
245
246A tracing session contains channel(s) which contains event(s). It is domain
247agnostic, meaning that channels and events can be enabled for the
248user-space tracer and/or the kernel tracer. It acts like a container
249aggregating multiple tracing sources.
250
251On creation, a \fB.lttngrc\fP file is created in your $HOME directory
252containing the current session name. If NAME is omitted, a session name is
253automatically created having this form: 'auto-yyyymmdd-hhmmss'.
254
255If no \fB\-o, \-\-output\fP is specified, the traces will be written in
256$HOME/lttng-traces.
257
258The $HOME environment variable can be overridden by defining the environment
259variable LTTNG_HOME. This is useful when the user running the commands has
260a non-writeable home directory.
261
262The session name MUST NOT contain the character '/'.
263
264.B OPTIONS:
265
266.TP
267.BR "\-h, \-\-help"
268Show summary of possible options and commands.
269.TP
270.BR "\-\-list-options"
271Simple listing of options
272.TP
273.BR "\-o, \-\-output PATH"
274Specify output path for traces
275.TP
276.BR "\-\-no-output"
277Traces will not be output
278.TP
279.BR "\-\-snapshot"
280Set the session in snapshot mode. Created in no-output mode and uses the
281URL, if one is specified, as the default snapshot output. Every channel will be set
282in overwrite mode and with mmap output (splice not supported).
283.TP
284.BR "\-\-live [USEC]"
285Set the session exclusively in live mode. The parameter is the delay in micro
286seconds before the data is flushed and streamed. The live mode allows you to
287stream the trace and view it while it's being recorded by any tracer. For that,
288you need a lttng-relayd and this session requires a network URL (\-U or
289\-C/\-D). If no USEC nor URL is provided, the default is to use a timer value
290set to 1000000 and the network URL set to net://127.0.0.1.
291
292To read a live session, you can use babeltrace(1) or the live streaming
293protocol in doc/live-reading-protocol.txt. Here is an example:
294
295.nf
296$ lttng-relayd -o /tmp/lttng
297$ lttng create --live 200000 -U net://localhost
298$ lttng enable-event -a --userspace
299$ lttng start
300.fi
301
302After the start, you'll be able to read the events while they are being
303recorded in /tmp/lttng.
304
305.TP
306.BR "\-U, \-\-set-url=URL"
307Set URL for the consumer output destination. It is persistent for the
308session lifetime. Redo the command to change it. This will set both data
309and control URL for network.
310.TP
311.BR "\-C, \-\-ctrl-url=URL"
312Set control path URL. (Must use -D also)
313.TP
314.BR "\-D, \-\-data-url=URL"
315Set data path URL. (Must use -C also)
316.PP
317Using these options, each API call can be controlled individually. For
318instance, \-C does not enable the consumer automatically. You'll need the \-e
319option for that.
320
321.B URL FORMAT:
322
323proto://[HOST|IP][:PORT1[:PORT2]][/TRACE_PATH]
324
325Supported protocols are (proto):
326.TP
327.BR "file://..."
328Local filesystem full path.
329
330.TP
331.BR "net://..."
332This will use the default network transport layer which is TCP for both
333control (PORT1) and data port (PORT2). The default ports are
334respectively 5342 and 5343. Note that net[6]:// is not yet supported.
335
336.TP
337.BR "tcp[6]://..."
338Can only be used with -C and -D together
339
340NOTE: IPv6 address MUST be enclosed in brackets '[]' (rfc2732)
341
342.B EXAMPLES:
343
344.nf
345# lttng create -U net://192.168.1.42
346.fi
347Uses TCP and default ports for the given destination.
348
349.nf
350# lttng create -U net6://[fe80::f66d:4ff:fe53:d220]
351.fi
352Uses TCP, default ports and IPv6.
353
354.nf
355# lttng create s1 -U net://myhost.com:3229
356.fi
357Create session s1 and set its consumer to myhost.com on port 3229 for control.
358.RE
359.PP
360
361.PP
362\fBdestroy\fP [NAME] [OPTIONS]
363.RS
364Teardown tracing session
365
366Free memory on the session daemon and tracer side. It's gone!
367
368If NAME is omitted, the session name is taken from the .lttngrc file.
369
370.B OPTIONS:
371
372.TP
373.BR "\-h, \-\-help"
374Show summary of possible options and commands.
375.TP
376.BR "\-a, \-\-all"
377Destroy all sessions
378.TP
379.BR "\-\-list-options"
380Simple listing of options
381.RE
382.PP
383
384.PP
385\fBenable-channel\fP NAME[,NAME2,...] (\-k | \-u) [OPTIONS]
386.RS
387Enable tracing channel
388
389To enable an event, you must enable both the event and the channel that
390contains it.
391
392If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
393file.
394
395Exactly one of \-k or -u must be specified.
396
397It is important to note that if a certain type of buffers is used, the session
398will be set with that type and all other subsequent channel needs to have the
399same type.
400
401Note that once the session has been started and enabled on the tracer side,
402it's not possible anymore to enable a new channel for that session.
403
404.B OPTIONS:
405
406.TP
407.BR "\-h, \-\-help"
408Show this help
409.TP
410.BR "\-\-list-options"
411Simple listing of options
412.TP
413.BR "\-s, \-\-session NAME"
414Apply on session name
415.TP
416.BR "\-k, \-\-kernel"
417Apply to the kernel tracer
418.TP
419.BR "\-u, \-\-userspace"
420Apply to the user-space tracer
421.TP
422.BR "\-\-discard"
423Discard event when subbuffers are full (default)
424.TP
425.BR "\-\-overwrite"
426Flight recorder mode: overwrites events when subbuffers are full. The
427number of subbuffer must be 2 or more.
428.TP
429.BR "\-\-subbuf-size SIZE"
430Subbuffer size in bytes {+k,+M,+G}.
431(default UST uid: 131072, UST pid: 4096, kernel: 262144, metadata: 4096)
432Rounded up to the next power of 2.
433
434The minimum subbuffer size, for each tracer, is the max value between
435the default above and the system page size. You can issue this command
436to get the current page size on your system: \fB$ getconf PAGE_SIZE\fP
437.TP
438.BR "\-\-num-subbuf NUM"
439Number of subbuffers. (default UST uid: 4, UST pid: 4, kernel: 4,
440metadata: 2) Rounded up to the next power of 2.
441.TP
442.BR "\-\-switch-timer USEC"
443Switch subbuffer timer interval in µsec.
444(default UST uid: 0, UST pid: 0, kernel: 0, metadata: 0)
445.TP
446.BR "\-\-read-timer USEC"
447Read timer interval in µsec.
448(default UST uid: 0, UST pid: 0, kernel: 200000, metadata: 0)
449.TP
450.BR "\-\-output TYPE"
451Channel output type. Possible values: mmap, splice
452(default UST uid: mmap, UST pid: mmap, kernel: splice, metadata: mmap)
453.TP
454.BR "\-\-buffers-uid"
455Use per UID buffer (\-u only). Buffers are shared between applications
456that have the same UID.
457.TP
458.BR "\-\-buffers-pid"
459Use per PID buffer (\-u only). Each application has its own buffers.
460.TP
461.BR "\-\-buffers-global"
462Use shared buffer for the whole system (\-k only)
463.TP
464.BR "\-C, \-\-tracefile-size SIZE"
465Maximum size of each tracefile within a stream (in bytes).
4660 means unlimited. (default: 0)
467Note: traces generated with this option may inaccurately report
468discarded events as of CTF 1.8.
469.TP
470.BR "\-W, \-\-tracefile-count COUNT"
471Used in conjunction with \-C option, this will limit the number of files
472created to the specified count. 0 means unlimited. (default: 0)
473
474.B EXAMPLES:
475
476.nf
477$ lttng enable-channel -k -C 4096 -W 32 chan1
478.fi
479For each stream, the maximum size of each trace file will be 4096 bytes and
480there will be a maximum of 32 different files. The file count is appended after
481the stream number as seen in the following example. The last trace file is
482smaller than 4096 since it was not completely filled.
483
484.nf
485 ~/lttng-traces/[...]/chan1_0_0 (4096)
486 ~/lttng-traces/[...]/chan1_0_1 (4096)
487 ~/lttng-traces/[...]/chan1_0_2 (3245)
488 ~/lttng-traces/[...]/chan1_1_0 (4096)
489 ...
490.fi
491
492.nf
493$ lttng enable-channel -k -C 4096
494.fi
495This will create trace files of 4096 bytes and will create new ones as long as
496there is data available.
497.RE
498.PP
499
500.PP
501\fBenable-event\fP NAME[,NAME2,...] [-k|-u] [OPTIONS]
502.RS
503Enable tracing event
504
505A tracing event is always assigned to a channel. If \fB\-c, \-\-channel\fP is
506omitted, a default channel named '\fBchannel0\fP' is created and the event is
507added to it. If \fB\-c, \-\-channel\fP is omitted, but a non-default
508channel already exists within the session, an error is returned. For the
509user-space tracer, using \fB\-a, \-\-all\fP is the same as using the
510wildcard "*".
511
512If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
513file.
514
515.B OPTIONS:
516
517.TP
518.BR "\-h, \-\-help"
519Show summary of possible options and commands.
520.TP
521.BR "\-\-list-options"
522Simple listing of options
523.TP
524.BR "\-s, \-\-session NAME"
525Apply on session name
526.TP
527.BR "\-c, \-\-channel NAME"
528Apply on channel name
529.TP
530.BR "\-a, \-\-all"
531Enable all tracepoints and syscalls. This actually enables a single
532wildcard event "*".
533.TP
534.BR "\-k, \-\-kernel"
535Apply for the kernel tracer
536.TP
537.BR "\-u, \-\-userspace"
538Apply for the user-space tracer
539.TP
540.BR "\-j, \-\-jul"
541Apply for Java application using Java Util Logging interface (JUL)
542.TP
543.BR "\-l, \-\-log4j"
544Apply for Java application using LOG4J
545.TP
546.BR "\-p, \-\-python"
547Apply for Python application using the logging module.
548.TP
549.BR "\-\-tracepoint"
550Tracepoint event (default). Userspace tracer supports wildcards at the end
551of string. Don't forget to quote to deal with bash expansion.
552e.g.:
553.nf
554 "*"
555 "app_component:na*"
556.fi
557.TP
558.BR "\-\-loglevel NAME"
559Tracepoint loglevel range from 0 to loglevel. Listed in the help (\-h).
560For the JUL domain, the loglevel ranges are detailed with the \-\-help
561option thus starting from SEVERE to FINEST.
562For the LOG4J domain, loglevels range from FATAL to TRACE which are also
563detailed in the help.
564For the Python domain, loglevels range from CRITICAL to DEBUG which are
565detailed in the help as well.
566.TP
567.BR "\-\-loglevel-only NAME"
568Tracepoint loglevel (only this loglevel).
569The loglevel or loglevel-only options should be combined with a
570tracepoint name or tracepoint wildcard.
571.TP
572.BR "\-\-probe (addr | symbol | symbol+offset)"
573Dynamic probe. Addr and offset can be octal (0NNN...), decimal (NNN...)
574or hexadecimal (0xNNN...)
575.TP
576.BR "\-\-function (addr | symbol | symbol+offset)"
577Dynamic function entry/return probe. Addr and offset can be octal
578(0NNN...), decimal (NNN...) or hexadecimal (0xNNN...)
579.TP
580.BR "\-\-syscall"
581System call event. Enabling syscalls tracing (kernel tracer), you will
582not be able to disable them with disable-event. This is a known
583limitation. You can disable the entire channel to do the trick. Also note
584that per-syscall selection is not supported yet. Use with "-a" to enable
585all syscalls.
586.TP
587.BR "\-\-filter 'expression'"
588Set a filter on a newly enabled event. Filter expression on event
589fields and context. The event will be recorded if the filter's
590expression evaluates to TRUE. Only specify on first activation of a
591given event within a session.
592Specifying a filter is only allowed when enabling events within a session before
593tracing is started. If the filter fails to link with the event
594within the traced domain, the event will be discarded.
595Filtering is currently only implemented for the user-space tracer.
596
597Expression examples:
598
599.nf
600 'intfield > 500 && intfield < 503'
601 '(strfield == "test" || intfield != 10) && intfield > 33'
602 'doublefield > 1.1 && intfield < 5.3'
603.fi
604
605Wildcards are allowed at the end of strings:
606 'seqfield1 == "te*"'
607In string literals, the escape character is a '\\'. Use '\\*' for
608the '*' character, and '\\\\' for the '\\' character sequence. Wildcard
609matches any sequence of characters, including an empty sub-string
610(matches 0 or more characters).
611
612Context information can be used for filtering. The examples below shows
613usage of context filtering on the process name (using a wildcard), process ID
614range, and unique thread ID. The process and thread IDs of
615running applications can be found under columns "PID" and "LWP" of the
616"ps -eLf" command.
617
618.nf
619 '$ctx.procname == "demo*"'
620 '$ctx.vpid >= 4433 && $ctx.vpid < 4455'
621 '$ctx.vtid == 1234'
622.fi
623
624Context information is available to all filters whether or not the add-context
625command has been used to add it to the event's channel, as long as the context
626field exists for that domain. For example, the filter examples given above will
627never fail to link: no add-context is required for the event's channel.
628
629.TP
630.BR "\-x, \-\-exclude LIST"
631Add exclusions to UST tracepoints:
632Events that match any of the items in the comma-separated LIST are not
633enabled, even if they match a wildcard definition of the event.
634
635This option is also applicable with the \fB\-a, \-\-all\fP option,
636in which case all UST tracepoints are enabled except the ones whose
637names match any of the items in LIST.
638.RE
639.PP
640
641.PP
642\fBdisable-channel\fP NAME[,NAME2,...] (\-k | \-u) [OPTIONS]
643.RS
644Disable tracing channel
645
646Disabling a channel disables the tracing of all of the channel's events. A channel
647can be re-enabled by calling \fBlttng enable-channel NAME\fP again.
648
649If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
650file.
651
652.B OPTIONS:
653
654.TP
655.BR "\-h, \-\-help"
656Show summary of possible options and commands.
657.TP
658.BR "\-\-list-options"
659Simple listing of options
660.TP
661.BR "\-s, \-\-session NAME"
662Apply on session name
663.TP
664.BR "\-k, \-\-kernel"
665Apply for the kernel tracer
666.TP
667.BR "\-u, \-\-userspace"
668Apply for the user-space tracer
669.RE
670.PP
671
672.PP
673\fBdisable-event\fP NAME[,NAME2,...] (\-k | \-u) [OPTIONS]
674.RS
675Disable tracing event
676
677The event, once disabled, can be re-enabled by calling \fBlttng enable-event
678NAME\fP again.
679
680If \fB\-s, \-\-session\fP is omitted, the session name is taken from the .lttngrc
681file.
682
683If \fB\-c, \-\-channel\fP is omitted, the default channel name is used.
684If \fB\-c, \-\-channel\fP is omitted, but a non-default channel already
685exists within the session, an error is returned.
686
687.B OPTIONS:
688
689.TP
690.BR "\-h, \-\-help"
691Show summary of possible options and commands.
692.TP
693.BR "\-\-list-options"
694Simple listing of options
695.TP
696.BR "\-s, \-\-session NAME"
697Apply on session name
698.TP
699.BR "\-c, \-\-channel NAME"
700Apply on channel name
701.TP
702.BR "\-a, \-\-all-events"
703Disable all events. This does NOT disable "*" but rather every known
704events of the session.
705.TP
706.BR "\-k, \-\-kernel"
707Apply for the kernel tracer
708.TP
709.BR "\-u, \-\-userspace"
710Apply for the user-space tracer
711.TP
712.BR "\-j, \-\-jul"
713Apply for Java application using Java Util Logging interface (JUL)
714.TP
715.BR "\-l, \-\-log4j"
716Apply for Java application using LOG4J
717.TP
718.BR "\-p, \-\-python"
719Apply for Python application using the logging module.
720.RE
721.PP
722
723.PP
724\fBlist\fP [OPTIONS] [SESSION [SESSION OPTIONS]]
725.RS
726List tracing session information.
727
728With no arguments, it will list available tracing session(s).
729
730With the session name, it will display the details of the session including
731the trace file path, the associated channels and their state (activated
732and deactivated), the activated events and more.
733
734With \-k alone, it will list all available kernel events (except the system
735calls events).
736With \-j alone, the available JUL event from registered application will be
737list. The event corresponds to the Logger name in the Java JUL application.
738With \-l alone, the available LOG4J event from registered application will be
739list. The event corresponds to the Logger name in the Java LOG4J application.
740With \-p alone, the available Python event from registered application will be
741list. The event corresponds to the Logger name in the Python application.
742With \-u alone, it will list all available user-space events from registered
743applications. Here is an example of 'lttng list \-u':
744
745.nf
746PID: 7448 - Name: /tmp/lttng-ust/tests/hello/.libs/lt-hello
747 ust_tests_hello:tptest_sighandler (type: tracepoint)
748 ust_tests_hello:tptest (type: tracepoint)
749.fi
750
751You can now enable any event listed by using the name :
752\fBust_tests_hello:tptest\fP.
753
754.B OPTIONS:
755
756.TP
757.BR "\-h, \-\-help"
758Show summary of possible options and commands.
759.TP
760.BR "\-\-list-options"
761Simple listing of options
762.TP
763.BR "\-k, \-\-kernel"
764Select kernel domain
765.TP
766.BR "\-u, \-\-userspace"
767Select user-space domain.
768.TP
769.BR "\-j, \-\-jul"
770Apply for Java application using JUL
771.TP
772.BR "\-l, \-\-log4j"
773Apply for Java application using LOG4J
774.TP
775.BR "\-p, \-\-python"
776Apply for Python application using the logging module.
777.TP
778.BR "\-f, \-\-fields"
779List event fields
780
781.PP
782.B SESSION OPTIONS:
783
784.TP
785.BR "\-c, \-\-channel NAME"
786List details of a channel
787.TP
788.BR "\-d, \-\-domain"
789List available domain(s)
790.RE
791.PP
792
793.PP
794\fBload\fP [OPTIONS] [NAME]
795.RS
796Load tracing session configuration
797
798If NAME is omitted, all session configurations found in both the user's session
799configuration directory (default: ~/.lttng/sessions/) and the system session
800configuration directory (default: /etc/lttng/sessions/) will be loaded. Note
801that the sessions in the user directory are loaded first and then the system
802wide directory are loaded.
803
804.B OPTIONS:
805
806.TP
807.BR "\-h, \-\-help"
808Show summary of possible options and commands.
809.TP
810.BR "\-a, \-\-all"
811Load all session configurations (default).
812.TP
813.BR "\-i, \-\-input-path PATH"
814Specify the input path for session configurations. This overrides the default
815session configuration directory.
816.TP
817.BR "\-f, -\-force"
818Overwrite current session configuration(s) if a session of the same name
819already exists.
820.RE
821.PP
822
823.PP
824\fBsave\fP [OPTIONS] [SESSION]
825.RS
826Save tracing session configuration
827
828If SESSION is omitted, all session configurations will be saved to individual
829\fB.lttng\fP files under the user's session configuration directory (default:
830~/.lttng/sessions/). The default session configuration file naming scheme is
831\fBSESSION.lttng\fP.
832
833For instance, a user in the tracing group saving a session from a root session
834daemon will save it in her/his user directory.
835
836.B OPTIONS:
837
838.TP
839.BR "\-h, \-\-help"
840Show summary of possible options and commands.
841.TP
842.BR "\-a, \-\-all"
843Save all session configurations (default).
844.TP
845.BR "\-o, \-\-output-path PATH"
846Specify the output path for saved sessions. This overrides the default session
847configuration directory.
848.TP
849.BR "\-f, -\-force"
850Overwrite session configuration file if session name clashes.
851.RE
852.PP
853
854.PP
855\fBset-session\fP NAME [OPTIONS]
856.RS
857Set current session name
858
859Will change the session name in the .lttngrc file.
860
861.B OPTIONS:
862
863.TP
864.BR "\-h, \-\-help"
865Show summary of possible options and commands.
866.TP
867.BR "\-\-list-options"
868Simple listing of options
869.RE
870.PP
871
872.PP
873\fBsnapshot\fP [OPTIONS] ACTION
874.RS
875Snapshot command for LTTng session.
876
877.B OPTIONS:
878
879.TP
880.BR "\-h, \-\-help"
881Show summary of possible options and commands.
882.TP
883.BR "\-\-list-options"
884Simple listing of options
885
886.PP
887.B ACTION:
888
889.TP
890\fBadd-output\fP [-m <SIZE>] [-s <NAME>] [-n <NAME>] <URL> | -C <URL> -D <URL>
891
892Setup and add a snapshot output for a session. Output is the destination
893where the snapshot will be sent. Only one output is permitted. To change it,
894you'll need to delete it and add back the new one.
895
896.TP
897\fBdel-output\fP ID | NAME [-s <NAME>]
898
899Delete an output for a session using the output's ID. You can either specify the
900output by name or use its ID as returned by the list-output command.
901
902.TP
903\fBlist-output\fP [-s <NAME>]
904
905List the output of a session. Attributes of the output are printed.
906
907.TP
908\fBrecord\fP [-m <SIZE>] [-s <NAME>] [-n <NAME>] [<URL> | -C <URL> -D <URL>]
909
910Snapshot a session's buffer(s) for all domains. If an URL is specified, it is
911used instead of a previously added output. Specifying only a name or/and a max
912size will override the current output values. For instance, you can record a
913snapshot with a custom maximum size or with a different name.
914
915.nf
916$ lttng snapshot add-output -n mysnapshot file:///data/snapshot
917[...]
918$ lttng snapshot record -n new_name_snapshot
919.fi
920
921The above will create a snapshot in /data/snapshot/new_name_snapshot* directory
922rather then in mysnapshot*/
923
924.PP
925.B DETAILED ACTION OPTIONS
926
927.TP
928.BR "\-s, \-\-session NAME"
929Apply to session name.
930.TP
931.BR "\-n, \-\-name NAME"
932Name of the snapshot's output.
933.TP
934.BR "\-m, \-\-max-size SIZE"
935Maximum size in bytes of the snapshot. The maximum size does not include the
936metadata file. Human readable format is accepted: {+k,+M,+G}. For instance,
937\-\-max-size 5M
938.TP
939.BR "\-C, \-\-ctrl-url URL"
940Set control path URL. (Must use -D also)
941.TP
942.BR "\-D, \-\-data-url URL"
943Set data path URL. (Must use -C also)
944.RE
945.PP
946
947.PP
948\fBstart\fP [NAME] [OPTIONS]
949.RS
950Start tracing
951
952It will start tracing for all tracers for a specific tracing session.
953If NAME is omitted, the session name is taken from the .lttngrc file.
954
955.B OPTIONS:
956
957.TP
958.BR "\-h, \-\-help"
959Show summary of possible options and commands.
960.TP
961.BR "\-\-list-options"
962Simple listing of options
963.RE
964.PP
965
966.PP
967\fBstop\fP [NAME] [OPTIONS]
968.RS
969Stop tracing
970
971It will stop tracing for all tracers for a specific tracing session. Before
972returning, the command checks for data availability meaning that it will wait
973until the trace is readable for the session. Use \-\-no-wait to avoid this
974behavior.
975
976If NAME is omitted, the session name is taken from the .lttngrc file.
977
978.B OPTIONS:
979
980.TP
981.BR "\-h, \-\-help"
982Show summary of possible options and commands.
983.TP
984.BR "\-\-list-options"
985Simple listing of options
986.TP
987.BR "\-\-no-wait"
988Don't wait for data availability.
989.RE
990.PP
991
992.PP
993\fBversion\fP
994.RS
995Show version information
996
997.B OPTIONS:
998
999.TP
1000.BR "\-h, \-\-help"
1001Show summary of possible options and commands.
1002.TP
1003.BR "\-\-list-options"
1004Simple listing of options
1005.RE
1006.PP
1007
1008.PP
1009\fBview\fP [SESSION_NAME] [OPTIONS]
1010.RS
1011View traces of a tracing session. By default, the babeltrace viewer
1012will be used for text viewing. If SESSION_NAME is omitted, the session
1013name is taken from the .lttngrc file.
1014
1015.B OPTIONS:
1016
1017.TP
1018.BR "\-h, \-\-help"
1019Show this help
1020.TP
1021.BR "\-\-list-options"
1022Simple listing of options
1023.TP
1024.BR "\-t, \-\-trace-path PATH"
1025Trace directory path for the viewer
1026.TP
1027.BR "\-e, \-\-viewer CMD"
1028Specify viewer and/or options to use This will completely override the
1029default viewers so please make sure to specify the full command. The
1030trace directory path of the session will be appended at the end to the
1031arguments
1032.RE
1033.PP
1034
1035.SH "JUL/LOG4J DOMAIN"
1036
1037This section explains the JUL and LOG4J domain where JUL stands for Java Util
1038Logging. You can use these by using the \fBliblttng-ust-<domain>-jni.so\fP from
1039the lttng-ust(3) project.
1040
1041The LTTng Java Agent uses JNI to link the UST tracer to the Java application
1042that uses the agent. Thus, it behaves similarly to the UST domain (\-u). When
1043enabling events, you enable a Logger name that will then be mapped to a default
1044UST tracepoint called \fBlttng_jul:<domain>_event\fP in the
1045\fBlttng_<domain>_channel\fP. Using the lttng-ctl API, any JUL/LOG4J events
1046must use the tracepoint event type (same as \-\-tracepoint).
1047
1048Because of the default immutable channel, the \fBenable-channel\fP command CAN
1049NOT be used with the JUL and LOG4J domain thus not having any options.
1050
1051Also, loglevels are supported. Use \fBlttng enable-event \-h\fP to list them.
1052Wildcards are NOT supported except the "*" meaning all events (same as \-a).
1053
1054Exactly like the UST domain, if the Java application has the same UID as you,
1055you can trace it. Same goes for the tracing group accessing root applications.
1056
1057Finally, you can list every Logger name that are available from registered
1058applications to the session daemon by using \fBlttng list \-j\fP or \fB\-l\fP.
1059
1060Here is an example on how to use the JUL domain.
1061
1062.nf
1063$ lttng list -j
1064[...]
1065$ lttng create aSession
1066$ lttng enable-event -s aSession -j MyCustomLoggerName
1067$ lttng start
1068.fi
1069
1070More information can be found in the lttng-ust documentation, see
1071java-util-logging.txt
1072.PP
1073
1074.SH "EXIT VALUES"
1075.PP
1076On success 0 is returned and a positive value on error. Value of 1 means a command
1077error, 2 an undefined command, 3 a fatal error and 4 a command warning meaning that
1078something went wrong during the command.
1079
1080Any other value above 10, please refer to
1081.BR "<lttng/lttng-error.h>"
1082for a detailed list or use lttng_strerror() to get a human readable string of
1083the error code.
1084.PP
1085
1086.SH "ENVIRONMENT VARIABLES"
1087
1088.PP
1089Note that all command line options override environment variables.
1090.PP
1091
1092.PP
1093.IP "LTTNG_SESSIOND_PATH"
1094Allows one to specify the full session daemon binary path to lttng command line
1095tool. You can also use \-\-sessiond-path option having the same effect.
1096.PP
1097
1098.PP
1099.IP "LTTNG_SESSION_CONFIG_XSD_PATH"
1100Set the path in which the \fBsession.xsd\fP session configuration schema may be
1101found.
1102.PP
1103
1104.SH "SEE ALSO"
1105.BR babeltrace(1),
1106.BR lttng-ust(3),
1107.BR lttng-sessiond(8),
1108.BR lttng-relayd(8),
1109
1110.SH "BUGS"
1111
1112.PP
1113If you encounter any issues or usability problem, please report it on our
1114mailing list <lttng-dev@lists.lttng.org> to help improve this project or
1115at https://bugs.lttng.org which is a bug tracker.
1116.PP
1117
1118.SH "CREDITS"
1119
1120.PP
1121lttng is distributed under the GNU General Public License version 2. See the file
1122COPYING for details.
1123.PP
1124A Web site is available at http://lttng.org for more information on the LTTng
1125project.
1126.PP
1127You can also find our git tree at http://git.lttng.org.
1128.PP
1129Mailing lists for support and development: <lttng-dev@lists.lttng.org>.
1130.PP
1131You can find us on IRC server irc.oftc.net (OFTC) in #lttng.
1132.PP
1133.SH "THANKS"
1134
1135.PP
1136Thanks to Yannick Brosseau without whom this project would never have been so
1137lean and mean! Also thanks to the Ericsson teams working on tracing which
1138helped us greatly with detailed bug reports and unusual test cases.
1139
1140Thanks to our beloved packager Alexandre Montplaisir-Goncalves (Ubuntu and PPA
1141maintainer) and Jon Bernard for our Debian packages.
1142
1143Special thanks to Michel Dagenais and the DORSAL laboratory at Polytechnique de
1144Montreal for the LTTng journey.
1145.PP
1146.SH "AUTHORS"
1147
1148.PP
1149lttng-tools was originally written by Mathieu Desnoyers, Julien Desfossez and
1150David Goulet. More people have since contributed to it. It is currently
1151maintained by Jérémie Galarneau <jeremie.galarneau@efficios.com>.
1152.PP
This page took 0.026731 seconds and 4 git commands to generate.