1 The LTTng Documentation
2 =======================
3 Philippe Proulx <pproulx@efficios.com>
7 include::../common/copyright.txt[]
10 include::../common/welcome.txt[]
13 include::../common/audience.txt[]
17 === What's in this documentation?
19 The LTTng Documentation is divided into the following sections:
21 * **<<nuts-and-bolts,Nuts and bolts>>** explains the
22 rudiments of software tracing and the rationale behind the
25 You can skip this section if you’re familiar with software tracing and
26 with the LTTng project.
28 * **<<installing-lttng,Installation>>** describes the steps to
29 install the LTTng packages on common Linux distributions and from
32 You can skip this section if you already properly installed LTTng on
35 * **<<getting-started,Quick start>>** is a concise guide to
36 getting started quickly with LTTng kernel and user space tracing.
38 We recommend this section if you're new to LTTng or to software tracing
41 You can skip this section if you're not new to LTTng.
43 * **<<core-concepts,Core concepts>>** explains the concepts at
46 It's a good idea to become familiar with the core concepts
47 before attempting to use the toolkit.
49 * **<<plumbing,Components of LTTng>>** describes the various components
50 of the LTTng machinery, like the daemons, the libraries, and the
51 command-line interface.
52 * **<<instrumenting,Instrumentation>>** shows different ways to
53 instrument user applications and the Linux kernel.
55 Instrumenting source code is essential to provide a meaningful
58 You can skip this section if you do not have a programming background.
60 * **<<controlling-tracing,Tracing control>>** is divided into topics
61 which demonstrate how to use the vast array of features that
62 LTTng{nbsp}{revision} offers.
63 * **<<reference,Reference>>** contains reference tables.
64 * **<<glossary,Glossary>>** is a specialized dictionary of terms related
65 to LTTng or to the field of software tracing.
68 include::../common/convention.txt[]
71 include::../common/acknowledgements.txt[]
75 == What's new in LTTng{nbsp}{revision}?
77 LTTng{nbsp}{revision} bears the name _Lafontaine_. This modern
78 https://en.wikipedia.org/wiki/saison[saison] from the
79 https://oshlag.com/[Oshlag] microbrewery is a refreshing--zesty--rice
80 beer with hints of fruit and spices. Some even say it makes for a great
81 https://en.wikipedia.org/wiki/Somaek[Somaek] when mixed with
82 Chamisul Soju, not that we've tried!
84 New features and changes in LTTng{nbsp}{revision}:
86 * Just like you can typically perform
87 https://en.wikipedia.org/wiki/Log_rotation[log rotation], you can
88 now <<session-rotation,_rotate_ a tracing session>>, that
89 is, according to man:lttng-rotate(1), archive the current trace
90 chunk (all the tracing session's trace data since the last rotation
91 or since its inception) so that LTTng does not manage it anymore.
93 Once LTTng archives a trace chunk, you are free to read it, modify it,
94 move it, or remove it.
96 You can rotate a tracing session immediately or set a rotation schedule
97 to automate rotations.
99 * When you <<enabling-disabling-events,create an event rule>>, the
100 filter expression syntax now supports the following new operators:
104 ** `<<` (bitwise left shift)
105 ** `>>` (bitwise right shift)
111 The syntax also supports array indexing with the usual square brackets:
114 regs[3][1] & 0xff7 == 0x240
117 There are peculiarities for both the new operators and the array
118 indexing brackets, like a custom precedence table and implicit casting.
119 See man:lttng-enable-event(1) to get all the details about the filter
122 * You can now dynamically instrument any application's or library's
123 function entry by symbol name thanks to the new
124 opt:lttng-enable-event(1):--userspace-probe option of
125 the `lttng enable-event` command:
129 $ lttng enable-event --kernel \
130 --userspace-probe=/usr/lib/libc.so.6:malloc libc_malloc
133 The option also supports tracing existing
134 https://www.sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps[SystemTap
135 Statically Defined Tracing] (USDT) probe (DTrace-style marker). For
136 example, given the following probe:
140 DTRACE_PROBE2("server", "accept-request", request_id, ip_addr);
143 You can trace this probe with:
146 $ lttng enable-event --kernel \
147 --userspace-probe=sdt:/path/to/server:server:accept-request \
148 server_accept_request
151 This feature makes use of Linux's
152 https://www.kernel.org/doc/Documentation/trace/uprobetracer.txt[uprobe]
153 mechanism, therefore you must use the `--userspace-probe`
154 instrumentation option with the opt:lttng-enable-event(1):--kernel
157 NOTE: As of LTTng{nbsp}{revision}, LTTng does not record function
158 parameters with the opt:lttng-enable-event(1):--userspace-probe option.
160 * Two new <<adding-context,context>> fields are available for Linux
161 kernel <<channel,channels>>:
164 ** `callstack-kernel`
168 Thanks to those, you can record the Linux kernel and user call stacks
169 when a kernel event occurs. For example:
173 $ lttng enable-event --kernel --syscall open
174 $ lttng add-context --kernel --type=callstack-kernel --type=callstack-user
177 When an man:open(2) system call occurs, LTTng attaches the kernel and
178 user call stacks to the recorded event.
180 NOTE: LTTng cannot always sample the user space call stack reliably.
181 For instance, LTTng cannot sample the call stack of user applications
182 and libraries compiled with the
183 https://gcc.gnu.org/onlinedocs/gcc/Optimize-Options.html[`-fomit-frame-pointer`]
184 option. In such a case, the tracing is not affected, but the sampled
185 user space call stack may only contain the user call stack's topmost
188 * User applications and libraries instrumented with
189 <<lttng-ust,LTTng-UST>> can now safely unload (man:dlclose(3)) a
191 <<building-tracepoint-providers-and-user-application,tracepoint
194 * The <<lttng-relayd,relay daemon>> is more efficient and presents fewer
195 connectivity issues, especially when a large number of targets send
196 trace data to a given relay daemon.
198 * LTTng-UST uses https://github.com/numactl/numactl[libnuma]
199 when available to allocate <<def-sub-buffer,sub-buffers>>, making them
201 https://en.wikipedia.org/wiki/Non-uniform_memory_access[NUMA] node.
203 This change makes the tracer more efficient on NUMA systems.
209 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
210 generation_ is a modern toolkit for tracing Linux systems and
211 applications. So your first question might be:
218 As the history of software engineering progressed and led to what
219 we now take for granted--complex, numerous and
220 interdependent software applications running in parallel on
221 sophisticated operating systems like Linux--the authors of such
222 components, software developers, began feeling a natural
223 urge to have tools that would ensure the robustness and good performance
224 of their masterpieces.
226 One major achievement in this field is, inarguably, the
227 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
228 an essential tool for developers to find and fix bugs. But even the best
229 debugger won't help make your software run faster, and nowadays, faster
230 software means either more work done by the same hardware, or cheaper
231 hardware for the same work.
233 A _profiler_ is often the tool of choice to identify performance
234 bottlenecks. Profiling is suitable to identify _where_ performance is
235 lost in a given software. The profiler outputs a profile, a statistical
236 summary of observed events, which you may use to discover which
237 functions took the most time to execute. However, a profiler won't
238 report _why_ some identified functions are the bottleneck. Bottlenecks
239 might only occur when specific conditions are met, conditions that are
240 sometimes impossible to capture by a statistical profiler, or impossible
241 to reproduce with an application altered by the overhead of an
242 event-based profiler. For a thorough investigation of software
243 performance issues, a history of execution is essential, with the
244 recorded values of variables and context fields you choose, and
245 with as little influence as possible on the instrumented software. This
246 is where tracing comes in handy.
248 _Tracing_ is a technique used to understand what goes on in a running
249 software system. The software used for tracing is called a _tracer_,
250 which is conceptually similar to a tape recorder. When recording,
251 specific instrumentation points placed in the software source code
252 generate events that are saved on a giant tape: a _trace_ file. You
253 can trace user applications and the operating system at the same time,
254 opening the possibility of resolving a wide range of problems that would
255 otherwise be extremely challenging.
257 Tracing is often compared to _logging_. However, tracers and loggers are
258 two different tools, serving two different purposes. Tracers are
259 designed to record much lower-level events that occur much more
260 frequently than log messages, often in the range of thousands per
261 second, with very little execution overhead. Logging is more appropriate
262 for a very high-level analysis of less frequent events: user accesses,
263 exceptional conditions (errors and warnings, for example), database
264 transactions, instant messaging communications, and such. Simply put,
265 logging is one of the many use cases that can be satisfied with tracing.
267 The list of recorded events inside a trace file can be read manually
268 like a log file for the maximum level of detail, but it is generally
269 much more interesting to perform application-specific analyses to
270 produce reduced statistics and graphs that are useful to resolve a
271 given problem. Trace viewers and analyzers are specialized tools
274 In the end, this is what LTTng is: a powerful, open source set of
275 tools to trace the Linux kernel and user applications at the same time.
276 LTTng is composed of several components actively maintained and
277 developed by its link:/community/#where[community].
280 [[lttng-alternatives]]
281 === Alternatives to noch:{LTTng}
283 Excluding proprietary solutions, a few competing software tracers
286 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
287 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
288 user scripts and is responsible for loading code into the
289 Linux kernel for further execution and collecting the outputted data.
290 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
291 subsystem in the Linux kernel in which a virtual machine can execute
292 programs passed from the user space to the kernel. You can attach
293 such programs to tracepoints and kprobes thanks to a system call, and
294 they can output data to the user space when executed thanks to
295 different mechanisms (pipe, VM register values, and eBPF maps, to name
297 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
298 is the de facto function tracer of the Linux kernel. Its user
299 interface is a set of special files in sysfs.
300 * https://perf.wiki.kernel.org/[perf] is
301 a performance analysis tool for Linux which supports hardware
302 performance counters, tracepoints, as well as other counters and
303 types of probes. perf's controlling utility is the cmd:perf command
305 * http://linux.die.net/man/1/strace[strace]
306 is a command-line utility which records system calls made by a
307 user process, as well as signal deliveries and changes of process
308 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
309 to fulfill its function.
310 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
311 analyze Linux kernel events. You write scripts, or _chisels_ in
312 sysdig's jargon, in Lua and sysdig executes them while it traces the
313 system or afterwards. sysdig's interface is the cmd:sysdig
314 command-line tool as well as the curses-based cmd:csysdig tool.
315 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
316 user space tracer which uses custom user scripts to produce plain text
317 traces. SystemTap converts the scripts to the C language, and then
318 compiles them as Linux kernel modules which are loaded to produce
319 trace data. SystemTap's primary user interface is the cmd:stap
322 The main distinctive features of LTTng is that it produces correlated
323 kernel and user space traces, as well as doing so with the lowest
324 overhead amongst other solutions. It produces trace files in the
325 http://diamon.org/ctf[CTF] format, a file format optimized
326 for the production and analyses of multi-gigabyte data.
328 LTTng is the result of more than 10{nbsp}years of active open source
329 development by a community of passionate developers.
330 LTTng{nbsp}{revision} is currently available on major desktop and server
333 The main interface for tracing control is a single command-line tool
334 named cmd:lttng. The latter can create several tracing sessions, enable
335 and disable events on the fly, filter events efficiently with custom
336 user expressions, start and stop tracing, and much more. LTTng can
337 record the traces on the file system or send them over the network, and
338 keep them totally or partially. You can view the traces once tracing
339 becomes inactive or in real-time.
341 <<installing-lttng,Install LTTng now>> and
342 <<getting-started,start tracing>>!
348 **LTTng** is a set of software <<plumbing,components>> which interact to
349 <<instrumenting,instrument>> the Linux kernel and user applications, and
350 to <<controlling-tracing,control tracing>> (start and stop
351 tracing, enable and disable event rules, and the rest). Those
352 components are bundled into the following packages:
355 Libraries and command-line interface to control tracing.
358 Linux kernel modules to instrument and trace the kernel.
361 Libraries and Java/Python packages to instrument and trace user
364 Most distributions mark the LTTng-modules and LTTng-UST packages as
365 optional when installing LTTng-tools (which is always required). In the
366 following sections, we always provide the steps to install all three,
369 * You only need to install LTTng-modules if you intend to trace the
371 * You only need to install LTTng-UST if you intend to trace user
376 As of 22 October 2019, LTTng{nbsp}{revision} is not available
377 as distribution packages, except for <<arch-linux,Arch Linux>>.
379 You can <<building-from-source,build LTTng{nbsp}{revision} from source>>
380 to install and use it.
387 LTTng-UST{nbsp}{revision} is available in Arch Linux's _community_
388 repository, while LTTng-tools{nbsp}{revision} and
389 LTTng-modules{nbsp}{revision} are available in the
390 https://aur.archlinux.org/[AUR].
392 To install LTTng{nbsp}{revision} on Arch Linux, using
393 https://github.com/actionless/pikaur[pikaur] for the AUR packages:
395 . Install the main LTTng{nbsp}{revision} packages:
400 # pacman -Sy lttng-ust
401 $ pikaur -Sy lttng-tools
402 $ pikaur -Sy lttng-modules
406 . **If you need to instrument and trace <<python-application,Python
407 applications>>**, install the LTTng-UST Python agent:
412 # pacman -Sy python-lttngust
413 # pacman -Sy python2-lttngust
418 [[building-from-source]]
419 === Build from source
421 To build and install LTTng{nbsp}{revision} from source:
423 . Using your distribution's package manager, or from source, install
424 the following dependencies of LTTng-tools and LTTng-UST:
427 * https://sourceforge.net/projects/libuuid/[libuuid]
428 * http://directory.fsf.org/wiki/Popt[popt]
429 * http://liburcu.org/[Userspace RCU]
430 * http://www.xmlsoft.org/[libxml2]
431 * **Optional**: https://github.com/numactl/numactl[numactl]
434 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
440 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.11.tar.bz2 &&
441 tar -xf lttng-modules-latest-2.11.tar.bz2 &&
442 cd lttng-modules-2.11.* &&
444 sudo make modules_install &&
449 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
455 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.11.tar.bz2 &&
456 tar -xf lttng-ust-latest-2.11.tar.bz2 &&
457 cd lttng-ust-2.11.* &&
465 Add `--disable-numa` to `./configure` if you don't have
466 https://github.com/numactl/numactl[numactl].
470 .Java and Python application tracing
472 If you need to instrument and trace <<java-application,Java
473 applications>>, pass the `--enable-java-agent-jul`,
474 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
475 `configure` script, depending on which Java logging framework you use.
477 If you need to instrument and trace <<python-application,Python
478 applications>>, pass the `--enable-python-agent` option to the
479 `configure` script. You can set the `PYTHON` environment variable to the
480 path to the Python interpreter for which to install the LTTng-UST Python
488 By default, LTTng-UST libraries are installed to
489 dir:{/usr/local/lib}, which is the de facto directory in which to
490 keep self-compiled and third-party libraries.
492 When <<building-tracepoint-providers-and-user-application,linking an
493 instrumented user application with `liblttng-ust`>>:
495 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
497 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
498 man:gcc(1), man:g++(1), or man:clang(1).
502 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
508 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.11.tar.bz2 &&
509 tar -xf lttng-tools-latest-2.11.tar.bz2 &&
510 cd lttng-tools-2.11.* &&
518 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
519 previous steps automatically for a given version of LTTng and confine
520 the installed files in a specific directory. This can be useful to test
521 LTTng without installing it on your system.
527 This is a short guide to get started quickly with LTTng kernel and user
530 Before you follow this guide, make sure to <<installing-lttng,install>>
533 This tutorial walks you through the steps to:
535 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
536 . <<tracing-your-own-user-application,Trace a user application>> written
538 . <<viewing-and-analyzing-your-traces,View and analyze the
542 [[tracing-the-linux-kernel]]
543 === Trace the Linux kernel
545 The following command lines start with the `#` prompt because you need
546 root privileges to trace the Linux kernel. You can also trace the kernel
547 as a regular user if your Unix user is a member of the
548 <<tracing-group,tracing group>>.
550 . Create a <<tracing-session,tracing session>> which writes its traces
551 to dir:{/tmp/my-kernel-trace}:
556 # lttng create my-kernel-session --output=/tmp/my-kernel-trace
560 . List the available kernel tracepoints and system calls:
565 # lttng list --kernel
566 # lttng list --kernel --syscall
570 . Create <<event,event rules>> which match the desired instrumentation
571 point names, for example the `sched_switch` and `sched_process_fork`
572 tracepoints, and the man:open(2) and man:close(2) system calls:
577 # lttng enable-event --kernel sched_switch,sched_process_fork
578 # lttng enable-event --kernel --syscall open,close
582 You can also create an event rule which matches _all_ the Linux kernel
583 tracepoints (this will generate a lot of data when tracing):
588 # lttng enable-event --kernel --all
592 . <<basic-tracing-session-control,Start tracing>>:
601 . Do some operation on your system for a few seconds. For example,
602 load a website, or list the files of a directory.
603 . <<creating-destroying-tracing-sessions,Destroy>> the current
613 The man:lttng-destroy(1) command does not destroy the trace data; it
614 only destroys the state of the tracing session.
616 The man:lttng-destroy(1) command also runs the man:lttng-stop(1) command
617 implicitly (see <<basic-tracing-session-control,Start and stop a tracing
618 session>>). You need to stop tracing to make LTTng flush the remaining
619 trace data and make the trace readable.
621 . For the sake of this example, make the recorded trace accessible to
627 # chown -R $(whoami) /tmp/my-kernel-trace
631 See <<viewing-and-analyzing-your-traces,View and analyze the
632 recorded events>> to view the recorded events.
635 [[tracing-your-own-user-application]]
636 === Trace a user application
638 This section steps you through a simple example to trace a
639 _Hello world_ program written in C.
641 To create the traceable user application:
643 . Create the tracepoint provider header file, which defines the
644 tracepoints and the events they can generate:
650 #undef TRACEPOINT_PROVIDER
651 #define TRACEPOINT_PROVIDER hello_world
653 #undef TRACEPOINT_INCLUDE
654 #define TRACEPOINT_INCLUDE "./hello-tp.h"
656 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
659 #include <lttng/tracepoint.h>
669 ctf_string(my_string_field, my_string_arg)
670 ctf_integer(int, my_integer_field, my_integer_arg)
674 #endif /* _HELLO_TP_H */
676 #include <lttng/tracepoint-event.h>
680 . Create the tracepoint provider package source file:
686 #define TRACEPOINT_CREATE_PROBES
687 #define TRACEPOINT_DEFINE
689 #include "hello-tp.h"
693 . Build the tracepoint provider package:
698 $ gcc -c -I. hello-tp.c
702 . Create the _Hello World_ application source file:
709 #include "hello-tp.h"
711 int main(int argc, char *argv[])
715 puts("Hello, World!\nPress Enter to continue...");
718 * The following getchar() call is only placed here for the purpose
719 * of this demonstration, to pause the application in order for
720 * you to have time to list its tracepoints. It is not
726 * A tracepoint() call.
728 * Arguments, as defined in hello-tp.h:
730 * 1. Tracepoint provider name (required)
731 * 2. Tracepoint name (required)
732 * 3. my_integer_arg (first user-defined argument)
733 * 4. my_string_arg (second user-defined argument)
735 * Notice the tracepoint provider and tracepoint names are
736 * NOT strings: they are in fact parts of variables that the
737 * macros in hello-tp.h create.
739 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
741 for (x = 0; x < argc; ++x) {
742 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
745 puts("Quitting now!");
746 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
753 . Build the application:
762 . Link the application with the tracepoint provider package,
763 `liblttng-ust`, and `libdl`:
768 $ gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
772 Here's the whole build process:
775 .User space tracing tutorial's build steps.
776 image::ust-flow.png[]
778 To trace the user application:
780 . Run the application with a few arguments:
785 $ ./hello world and beyond
794 Press Enter to continue...
798 . Start an LTTng <<lttng-sessiond,session daemon>>:
803 $ lttng-sessiond --daemonize
807 Note that a session daemon might already be running, for example as
808 a service that the distribution's service manager started.
810 . List the available user space tracepoints:
815 $ lttng list --userspace
819 You see the `hello_world:my_first_tracepoint` tracepoint listed
820 under the `./hello` process.
822 . Create a <<tracing-session,tracing session>>:
827 $ lttng create my-user-space-session
831 . Create an <<event,event rule>> which matches the
832 `hello_world:my_first_tracepoint` event name:
837 $ lttng enable-event --userspace hello_world:my_first_tracepoint
841 . <<basic-tracing-session-control,Start tracing>>:
850 . Go back to the running `hello` application and press Enter. The
851 program executes all `tracepoint()` instrumentation points and exits.
852 . <<creating-destroying-tracing-sessions,Destroy>> the current
862 The man:lttng-destroy(1) command does not destroy the trace data; it
863 only destroys the state of the tracing session.
865 The man:lttng-destroy(1) command also runs the man:lttng-stop(1) command
866 implicitly (see <<basic-tracing-session-control,Start and stop a tracing
867 session>>). You need to stop tracing to make LTTng flush the remaining
868 trace data and make the trace readable.
870 By default, LTTng saves the traces in
871 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
872 where +__name__+ is the tracing session name. The
873 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
875 See <<viewing-and-analyzing-your-traces,View and analyze the
876 recorded events>> to view the recorded events.
879 [[viewing-and-analyzing-your-traces]]
880 === View and analyze the recorded events
882 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
883 kernel>> and <<tracing-your-own-user-application,Trace a user
884 application>> tutorials, you can inspect the recorded events.
886 There are many tools you can use to read LTTng traces:
888 * **cmd:babeltrace** is a command-line utility which converts trace
889 formats; it supports the format that LTTng produces, CTF, as well as a
890 basic text output which can be ++grep++ed. The cmd:babeltrace command
891 is part of the http://diamon.org/babeltrace[Babeltrace] project.
892 * Babeltrace also includes
893 **https://www.python.org/[Python{nbsp}3] bindings** so
894 that you can easily open and read an LTTng trace with your own script,
895 benefiting from the power of Python.
896 * http://tracecompass.org/[**Trace Compass**]
897 is a graphical user interface for viewing and analyzing any type of
898 logs or traces, including LTTng's.
899 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
900 project which includes many high-level analyses of LTTng kernel
901 traces, like scheduling statistics, interrupt frequency distribution,
902 top CPU usage, and more.
904 NOTE: This section assumes that LTTng saved the traces it recorded
905 during the previous tutorials to their default location, in the
906 dir:{$LTTNG_HOME/lttng-traces} directory. The env:LTTNG_HOME
907 environment variable defaults to `$HOME` if not set.
910 [[viewing-and-analyzing-your-traces-bt]]
911 ==== Use the cmd:babeltrace command-line tool
913 The simplest way to list all the recorded events of a trace is to pass
914 its path to cmd:babeltrace with no options:
918 $ babeltrace ~/lttng-traces/my-user-space-session*
921 cmd:babeltrace finds all traces recursively within the given path and
922 prints all their events, merging them in chronological order.
924 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
929 $ babeltrace /tmp/my-kernel-trace | grep _switch
932 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
933 count the recorded events:
937 $ babeltrace /tmp/my-kernel-trace | grep _open | wc --lines
941 [[viewing-and-analyzing-your-traces-bt-python]]
942 ==== Use the Babeltrace{nbsp}1 Python bindings
944 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
945 is useful to isolate events by simple matching using man:grep(1) and
946 similar utilities. However, more elaborate filters, such as keeping only
947 event records with a field value falling within a specific range, are
948 not trivial to write using a shell. Moreover, reductions and even the
949 most basic computations involving multiple event records are virtually
950 impossible to implement.
952 Fortunately, Babeltrace{nbsp}1 ships with Python{nbsp}3 bindings which
953 makes it easy to read the event records of an LTTng trace sequentially
954 and compute the desired information.
956 The following script accepts an LTTng Linux kernel trace path as its
957 first argument and prints the short names of the top five running
958 processes on CPU{nbsp}0 during the whole trace:
963 from collections import Counter
969 if len(sys.argv) != 2:
970 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
971 print(msg, file=sys.stderr)
974 # A trace collection contains one or more traces
975 col = babeltrace.TraceCollection()
977 # Add the trace provided by the user (LTTng traces always have
979 if col.add_trace(sys.argv[1], 'ctf') is None:
980 raise RuntimeError('Cannot add trace')
982 # This counter dict contains execution times:
984 # task command name -> total execution time (ns)
985 exec_times = Counter()
987 # This contains the last `sched_switch` timestamp
991 for event in col.events:
992 # Keep only `sched_switch` events
993 if event.name != 'sched_switch':
996 # Keep only events which happened on CPU 0
997 if event['cpu_id'] != 0:
1001 cur_ts = event.timestamp
1007 # Previous task command (short) name
1008 prev_comm = event['prev_comm']
1010 # Initialize entry in our dict if not yet done
1011 if prev_comm not in exec_times:
1012 exec_times[prev_comm] = 0
1014 # Compute previous command execution time
1015 diff = cur_ts - last_ts
1017 # Update execution time of this command
1018 exec_times[prev_comm] += diff
1020 # Update last timestamp
1024 for name, ns in exec_times.most_common(5):
1026 print('{:20}{} s'.format(name, s))
1031 if __name__ == '__main__':
1032 sys.exit(0 if top5proc() else 1)
1039 $ python3 top5proc.py /tmp/my-kernel-trace/kernel
1045 swapper/0 48.607245889 s
1046 chromium 7.192738188 s
1047 pavucontrol 0.709894415 s
1048 Compositor 0.660867933 s
1049 Xorg.bin 0.616753786 s
1052 Note that `swapper/0` is the "idle" process of CPU{nbsp}0 on Linux;
1053 since we weren't using the CPU that much when tracing, its first
1054 position in the list makes sense.
1058 == [[understanding-lttng]]Core concepts
1060 From a user's perspective, the LTTng system is built on a few concepts,
1061 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1062 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1063 Understanding how those objects relate to eachother is key in mastering
1066 The core concepts are:
1068 * <<tracing-session,Tracing session>>
1069 * <<domain,Tracing domain>>
1070 * <<channel,Channel and ring buffer>>
1071 * <<"event","Instrumentation point, event rule, event, and event record">>
1077 A _tracing session_ is a stateful dialogue between you and
1078 a <<lttng-sessiond,session daemon>>. You can
1079 <<creating-destroying-tracing-sessions,create a new tracing
1080 session>> with the `lttng create` command.
1082 Anything that you do when you control LTTng tracers happens within a
1083 tracing session. In particular, a tracing session:
1086 * Has its own set of trace files.
1087 * Has its own state of activity (started or stopped).
1088 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1090 * Has its own <<channel,channels>> to which are associated their own
1091 <<event,event rules>>.
1094 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1095 image::concepts.png[]
1097 Those attributes and objects are completely isolated between different
1100 A tracing session is analogous to a cash machine session:
1101 the operations you do on the banking system through the cash machine do
1102 not alter the data of other users of the same system. In the case of
1103 the cash machine, a session lasts as long as your bank card is inside.
1104 In the case of LTTng, a tracing session lasts from the `lttng create`
1105 command to the `lttng destroy` command.
1108 .Each Unix user has its own set of tracing sessions.
1109 image::many-sessions.png[]
1112 [[tracing-session-mode]]
1113 ==== Tracing session mode
1115 LTTng can send the generated trace data to different locations. The
1116 _tracing session mode_ dictates where to send it. The following modes
1117 are available in LTTng{nbsp}{revision}:
1120 LTTng writes the traces to the file system of the machine it traces
1123 Network streaming mode::
1124 LTTng sends the traces over the network to a
1125 <<lttng-relayd,relay daemon>> running on a remote system.
1128 LTTng does not write the traces by default. Instead, you can request
1129 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1130 tracing session's current sub-buffers, and to write it to the
1131 target's file system or to send it over the network to a
1132 <<lttng-relayd,relay daemon>> running on a remote system.
1135 This mode is similar to the network streaming mode, but a live
1136 trace viewer can connect to the distant relay daemon to
1137 <<lttng-live,view event records as LTTng generates them>>.
1143 A _tracing domain_ is a namespace for event sources. A tracing domain
1144 has its own properties and features.
1146 There are currently five available tracing domains:
1150 * `java.util.logging` (JUL)
1154 You must specify a tracing domain when using some commands to avoid
1155 ambiguity. For example, since all the domains support named tracepoints
1156 as event sources (instrumentation points that you manually insert in the
1157 source code), you need to specify a tracing domain when
1158 <<enabling-disabling-events,creating an event rule>> because all the
1159 tracing domains could have tracepoints with the same names.
1161 You can create <<channel,channels>> in the Linux kernel and user space
1162 tracing domains. The other tracing domains have a single default
1167 === Channel and ring buffer
1169 A _channel_ is an object which is responsible for a set of ring buffers.
1170 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1171 tracer emits an event, it can record it to one or more
1172 sub-buffers. The attributes of a channel determine what to do when
1173 there's no space left for a new event record because all sub-buffers
1174 are full, where to send a full sub-buffer, and other behaviours.
1176 A channel is always associated to a <<domain,tracing domain>>. The
1177 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1178 a default channel which you cannot configure.
1180 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1181 an event, it records it to the sub-buffers of all
1182 the enabled channels with a satisfied event rule, as long as those
1183 channels are part of active <<tracing-session,tracing sessions>>.
1186 [[channel-buffering-schemes]]
1187 ==== Per-user vs. per-process buffering schemes
1189 A channel has at least one ring buffer _per CPU_. LTTng always
1190 records an event to the ring buffer associated to the CPU on which it
1193 Two _buffering schemes_ are available when you
1194 <<enabling-disabling-channels,create a channel>> in the
1195 user space <<domain,tracing domain>>:
1197 Per-user buffering::
1198 Allocate one set of ring buffers--one per CPU--shared by all the
1199 instrumented processes of each Unix user.
1203 .Per-user buffering scheme.
1204 image::per-user-buffering.png[]
1207 Per-process buffering::
1208 Allocate one set of ring buffers--one per CPU--for each
1209 instrumented process.
1213 .Per-process buffering scheme.
1214 image::per-process-buffering.png[]
1217 The per-process buffering scheme tends to consume more memory than the
1218 per-user option because systems generally have more instrumented
1219 processes than Unix users running instrumented processes. However, the
1220 per-process buffering scheme ensures that one process having a high
1221 event throughput won't fill all the shared sub-buffers of the same
1224 The Linux kernel tracing domain has only one available buffering scheme
1225 which is to allocate a single set of ring buffers for the whole system.
1226 This scheme is similar to the per-user option, but with a single, global
1227 user "running" the kernel.
1230 [[channel-overwrite-mode-vs-discard-mode]]
1231 ==== Overwrite vs. discard event record loss modes
1233 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1234 arc in the following animations) of a specific channel's ring buffer.
1235 When there's no space left in a sub-buffer, the tracer marks it as
1236 consumable (red) and another, empty sub-buffer starts receiving the
1237 following event records. A <<lttng-consumerd,consumer daemon>>
1238 eventually consumes the marked sub-buffer (returns to white).
1241 [role="docsvg-channel-subbuf-anim"]
1246 In an ideal world, sub-buffers are consumed faster than they are filled,
1247 as it is the case in the previous animation. In the real world,
1248 however, all sub-buffers can be full at some point, leaving no space to
1249 record the following events.
1251 By default, LTTng-modules and LTTng-UST are _non-blocking_ tracers: when
1252 no empty sub-buffer is available, it is acceptable to lose event records
1253 when the alternative would be to cause substantial delays in the
1254 instrumented application's execution. LTTng privileges performance over
1255 integrity; it aims at perturbing the target system as little as possible
1256 in order to make tracing of subtle race conditions and rare interrupt
1259 Since LTTng{nbsp}2.10, the LTTng user space tracer, LTTng-UST, supports
1260 a _blocking mode_. See the <<blocking-timeout-example,blocking timeout
1261 example>> to learn how to use the blocking mode.
1263 When it comes to losing event records because no empty sub-buffer is
1264 available, or because the <<opt-blocking-timeout,blocking timeout>> is
1265 reached, the channel's _event record loss mode_ determines what to do.
1266 The available event record loss modes are:
1269 Drop the newest event records until the tracer releases a sub-buffer.
1271 This is the only available mode when you specify a
1272 <<opt-blocking-timeout,blocking timeout>>.
1275 Clear the sub-buffer containing the oldest event records and start
1276 writing the newest event records there.
1278 This mode is sometimes called _flight recorder mode_ because it's
1280 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1281 always keep a fixed amount of the latest data.
1283 Which mechanism you should choose depends on your context: prioritize
1284 the newest or the oldest event records in the ring buffer?
1286 Beware that, in overwrite mode, the tracer abandons a _whole sub-buffer_
1287 as soon as a there's no space left for a new event record, whereas in
1288 discard mode, the tracer only discards the event record that doesn't
1291 In discard mode, LTTng increments a count of lost event records when an
1292 event record is lost and saves this count to the trace. In overwrite
1293 mode, since LTTng{nbsp}2.8, LTTng increments a count of lost sub-buffers
1294 when a sub-buffer is lost and saves this count to the trace. In this
1295 mode, LTTng does not write to the trace the exact number of lost event
1296 records in those lost sub-buffers. Trace analyses can use the trace's
1297 saved discarded event record and sub-buffer counts to decide whether or
1298 not to perform the analyses even if trace data is known to be missing.
1300 There are a few ways to decrease your probability of losing event
1302 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1303 how you can fine-tune the sub-buffer count and size of a channel to
1304 virtually stop losing event records, though at the cost of greater
1308 [[channel-subbuf-size-vs-subbuf-count]]
1309 ==== Sub-buffer count and size
1311 When you <<enabling-disabling-channels,create a channel>>, you can
1312 set its number of sub-buffers and their size.
1314 Note that there is noticeable CPU overhead introduced when
1315 switching sub-buffers (marking a full one as consumable and switching
1316 to an empty one for the following events to be recorded). Knowing this,
1317 the following list presents a few practical situations along with how
1318 to configure the sub-buffer count and size for them:
1320 * **High event throughput**: In general, prefer bigger sub-buffers to
1321 lower the risk of losing event records.
1323 Having bigger sub-buffers also ensures a lower
1324 <<channel-switch-timer,sub-buffer switching frequency>>.
1326 The number of sub-buffers is only meaningful if you create the channel
1327 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1328 other sub-buffers are left unaltered.
1330 * **Low event throughput**: In general, prefer smaller sub-buffers
1331 since the risk of losing event records is low.
1333 Because events occur less frequently, the sub-buffer switching frequency
1334 should remain low and thus the tracer's overhead should not be a
1337 * **Low memory system**: If your target system has a low memory
1338 limit, prefer fewer first, then smaller sub-buffers.
1340 Even if the system is limited in memory, you want to keep the
1341 sub-buffers as big as possible to avoid a high sub-buffer switching
1344 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1345 which means event data is very compact. For example, the average
1346 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1347 sub-buffer size of 1{nbsp}MiB is considered big.
1349 The previous situations highlight the major trade-off between a few big
1350 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1351 frequency vs. how much data is lost in overwrite mode. Assuming a
1352 constant event throughput and using the overwrite mode, the two
1353 following configurations have the same ring buffer total size:
1356 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1361 * **Two sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1362 switching frequency, but if a sub-buffer overwrite happens, half of
1363 the event records so far (4{nbsp}MiB) are definitely lost.
1364 * **Eight sub-buffers of 1{nbsp}MiB each**: Expect four times the tracer's
1365 overhead as the previous configuration, but if a sub-buffer
1366 overwrite happens, only the eighth of event records so far are
1369 In discard mode, the sub-buffers count parameter is pointless: use two
1370 sub-buffers and set their size according to the requirements of your
1374 [[channel-switch-timer]]
1375 ==== Switch timer period
1377 The _switch timer period_ is an important configurable attribute of
1378 a channel to ensure periodic sub-buffer flushing.
1380 When the _switch timer_ expires, a sub-buffer switch happens. You can
1381 set the switch timer period attribute when you
1382 <<enabling-disabling-channels,create a channel>> to ensure that LTTng
1383 consumes and commits trace data to trace files or to a distant relay
1384 daemon periodically in case of a low event throughput.
1387 [role="docsvg-channel-switch-timer"]
1392 This attribute is also convenient when you use big sub-buffers to cope
1393 with a sporadic high event throughput, even if the throughput is
1397 [[channel-read-timer]]
1398 ==== Read timer period
1400 By default, the LTTng tracers use a notification mechanism to signal a
1401 full sub-buffer so that a consumer daemon can consume it. When such
1402 notifications must be avoided, for example in real-time applications,
1403 you can use the channel's _read timer_ instead. When the read timer
1404 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1405 consumable sub-buffers.
1408 [[tracefile-rotation]]
1409 ==== Trace file count and size
1411 By default, trace files can grow as large as needed. You can set the
1412 maximum size of each trace file that a channel writes when you
1413 <<enabling-disabling-channels,create a channel>>. When the size of
1414 a trace file reaches the channel's fixed maximum size, LTTng creates
1415 another file to contain the next event records. LTTng appends a file
1416 count to each trace file name in this case.
1418 If you set the trace file size attribute when you create a channel, the
1419 maximum number of trace files that LTTng creates is _unlimited_ by
1420 default. To limit them, you can also set a maximum number of trace
1421 files. When the number of trace files reaches the channel's fixed
1422 maximum count, the oldest trace file is overwritten. This mechanism is
1423 called _trace file rotation_.
1427 Even if you don't limit the trace file count, you cannot assume that
1428 LTTng doesn't manage any trace file.
1430 In other words, there is no safe way to know if LTTng still holds a
1431 given trace file open with the trace file rotation feature.
1433 The only way to obtain an unmanaged, self-contained LTTng trace before
1434 you <<creating-destroying-tracing-sessions,destroy>> the tracing session
1435 is with the <<session-rotation,tracing session rotation>> feature
1436 (available since LTTng{nbsp}2.11).
1441 === Instrumentation point, event rule, event, and event record
1443 An _event rule_ is a set of conditions which must be **all** satisfied
1444 for LTTng to record an occuring event.
1446 You set the conditions when you <<enabling-disabling-events,create
1449 You always attach an event rule to a <<channel,channel>> when you create
1452 When an event passes the conditions of an event rule, LTTng records it
1453 in one of the attached channel's sub-buffers.
1455 The available conditions, as of LTTng{nbsp}{revision}, are:
1457 * The event rule _is enabled_.
1458 * The instrumentation point's type _is{nbsp}T_.
1459 * The instrumentation point's name (sometimes called _event name_)
1460 _matches{nbsp}N_, but _is not{nbsp}E_.
1461 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1462 _is exactly{nbsp}L_.
1463 * The fields of the event's payload _satisfy_ a filter
1464 expression{nbsp}__F__.
1466 As you can see, all the conditions but the dynamic filter are related to
1467 the event rule's status or to the instrumentation point, not to the
1468 occurring events. This is why, without a filter, checking if an event
1469 passes an event rule is not a dynamic task: when you create or modify an
1470 event rule, all the tracers of its tracing domain enable or disable the
1471 instrumentation points themselves once. This is possible because the
1472 attributes of an instrumentation point (type, name, and log level) are
1473 defined statically. In other words, without a dynamic filter, the tracer
1474 _does not evaluate_ the arguments of an instrumentation point unless it
1475 matches an enabled event rule.
1477 Note that, for LTTng to record an event, the <<channel,channel>> to
1478 which a matching event rule is attached must also be enabled, and the
1479 <<tracing-session,tracing session>> owning this channel must be active
1483 .Logical path from an instrumentation point to an event record.
1484 image::event-rule.png[]
1486 .Event, event record, or event rule?
1488 With so many similar terms, it's easy to get confused.
1490 An **event** is the consequence of the execution of an _instrumentation
1491 point_, like a tracepoint that you manually place in some source code,
1492 or a Linux kernel kprobe. An event is said to _occur_ at a specific
1493 time. Different actions can be taken upon the occurrence of an event,
1494 like record the event's payload to a buffer.
1496 An **event record** is the representation of an event in a sub-buffer. A
1497 tracer is responsible for capturing the payload of an event, current
1498 context variables, the event's ID, and the event's timestamp. LTTng
1499 can append this sub-buffer to a trace file.
1501 An **event rule** is a set of conditions which must _all_ be satisfied
1502 for LTTng to record an occuring event. Events still occur without
1503 satisfying event rules, but LTTng does not record them.
1508 == Components of noch:{LTTng}
1510 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1511 to call LTTng a simple _tool_ since it is composed of multiple
1512 interacting components. This section describes those components,
1513 explains their respective roles, and shows how they connect together to
1514 form the LTTng ecosystem.
1516 The following diagram shows how the most important components of LTTng
1517 interact with user applications, the Linux kernel, and you:
1520 .Control and trace data paths between LTTng components.
1521 image::plumbing.png[]
1523 The LTTng project incorporates:
1525 * **LTTng-tools**: Libraries and command-line interface to
1526 control tracing sessions.
1527 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1528 ** <<lttng-consumerd,Consumer daemon>> (cmd:lttng-consumerd).
1529 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1530 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1531 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1532 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1534 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1535 headers to instrument and trace any native user application.
1536 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1537 *** `liblttng-ust-libc-wrapper`
1538 *** `liblttng-ust-pthread-wrapper`
1539 *** `liblttng-ust-cyg-profile`
1540 *** `liblttng-ust-cyg-profile-fast`
1541 *** `liblttng-ust-dl`
1542 ** User space tracepoint provider source files generator command-line
1543 tool (man:lttng-gen-tp(1)).
1544 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1545 Java applications using `java.util.logging` or
1546 Apache log4j{nbsp}1.2 logging.
1547 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1548 Python applications using the standard `logging` package.
1549 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1551 ** LTTng kernel tracer module.
1552 ** Tracing ring buffer kernel modules.
1553 ** Probe kernel modules.
1554 ** LTTng logger kernel module.
1558 === Tracing control command-line interface
1561 .The tracing control command-line interface.
1562 image::plumbing-lttng-cli.png[]
1564 The _man:lttng(1) command-line tool_ is the standard user interface to
1565 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1566 is part of LTTng-tools.
1568 The cmd:lttng tool is linked with
1569 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1570 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1572 The cmd:lttng tool has a Git-like interface:
1576 $ lttng <GENERAL OPTIONS> <COMMAND> <COMMAND OPTIONS>
1579 The <<controlling-tracing,Tracing control>> section explores the
1580 available features of LTTng using the cmd:lttng tool.
1583 [[liblttng-ctl-lttng]]
1584 === Tracing control library
1587 .The tracing control library.
1588 image::plumbing-liblttng-ctl.png[]
1590 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1591 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1592 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1594 The <<lttng-cli,cmd:lttng command-line tool>>
1595 is linked with `liblttng-ctl`.
1597 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1602 #include <lttng/lttng.h>
1605 Some objects are referenced by name (C string), such as tracing
1606 sessions, but most of them require to create a handle first using
1607 `lttng_create_handle()`.
1609 As of LTTng{nbsp}{revision}, the best available developer documentation for
1610 `liblttng-ctl` is its installed header files. Every function and structure is
1611 thoroughly documented.
1615 === User space tracing library
1618 .The user space tracing library.
1619 image::plumbing-liblttng-ust.png[]
1621 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1622 is the LTTng user space tracer. It receives commands from a
1623 <<lttng-sessiond,session daemon>>, for example to
1624 enable and disable specific instrumentation points, and writes event
1625 records to ring buffers shared with a
1626 <<lttng-consumerd,consumer daemon>>.
1627 `liblttng-ust` is part of LTTng-UST.
1629 Public C header files are installed beside `liblttng-ust` to
1630 instrument any <<c-application,C or $$C++$$ application>>.
1632 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1633 packages, use their own library providing tracepoints which is
1634 linked with `liblttng-ust`.
1636 An application or library does not have to initialize `liblttng-ust`
1637 manually: its constructor does the necessary tasks to properly register
1638 to a session daemon. The initialization phase also enables the
1639 instrumentation points matching the <<event,event rules>> that you
1643 [[lttng-ust-agents]]
1644 === User space tracing agents
1647 .The user space tracing agents.
1648 image::plumbing-lttng-ust-agents.png[]
1650 The _LTTng-UST Java and Python agents_ are regular Java and Python
1651 packages which add LTTng tracing capabilities to the
1652 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1654 In the case of Java, the
1655 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1656 core logging facilities] and
1657 https://logging.apache.org/log4j/1.2/[Apache log4j{nbsp}1.2] are supported.
1658 Note that Apache Log4{nbsp}2 is not supported.
1660 In the case of Python, the standard
1661 https://docs.python.org/3/library/logging.html[`logging`] package
1662 is supported. Both Python{nbsp}2 and Python{nbsp}3 modules can import the
1663 LTTng-UST Python agent package.
1665 The applications using the LTTng-UST agents are in the
1666 `java.util.logging` (JUL),
1667 log4j, and Python <<domain,tracing domains>>.
1669 Both agents use the same mechanism to trace the log statements. When an
1670 agent initializes, it creates a log handler that attaches to the root
1671 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1672 When the application executes a log statement, the root logger passes it
1673 to the agent's log handler. The agent's log handler calls a native
1674 function in a tracepoint provider package shared library linked with
1675 <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1676 other fields, like its logger name and its log level. This native
1677 function contains a user space instrumentation point, hence tracing the
1680 The log level condition of an
1681 <<event,event rule>> is considered when tracing
1682 a Java or a Python application, and it's compatible with the standard
1683 JUL, log4j, and Python log levels.
1687 === LTTng kernel modules
1690 .The LTTng kernel modules.
1691 image::plumbing-lttng-modules.png[]
1693 The _LTTng kernel modules_ are a set of Linux kernel modules
1694 which implement the kernel tracer of the LTTng project. The LTTng
1695 kernel modules are part of LTTng-modules.
1697 The LTTng kernel modules include:
1699 * A set of _probe_ modules.
1701 Each module attaches to a specific subsystem
1702 of the Linux kernel using its tracepoint instrument points. There are
1703 also modules to attach to the entry and return points of the Linux
1704 system call functions.
1706 * _Ring buffer_ modules.
1708 A ring buffer implementation is provided as kernel modules. The LTTng
1709 kernel tracer writes to the ring buffer; a
1710 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1712 * The _LTTng kernel tracer_ module.
1713 * The _LTTng logger_ module.
1715 The LTTng logger module implements the special path:{/proc/lttng-logger}
1716 file so that any executable can generate LTTng events by opening and
1717 writing to this file.
1719 See <<proc-lttng-logger-abi,LTTng logger>>.
1721 Generally, you do not have to load the LTTng kernel modules manually
1722 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1723 daemon>> loads the necessary modules when starting. If you have extra
1724 probe modules, you can specify to load them to the session daemon on
1727 The LTTng kernel modules are installed in
1728 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1729 the kernel release (see `uname --kernel-release`).
1736 .The session daemon.
1737 image::plumbing-sessiond.png[]
1739 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1740 managing tracing sessions and for controlling the various components of
1741 LTTng. The session daemon is part of LTTng-tools.
1743 The session daemon sends control requests to and receives control
1746 * The <<lttng-ust,user space tracing library>>.
1748 Any instance of the user space tracing library first registers to
1749 a session daemon. Then, the session daemon can send requests to
1750 this instance, such as:
1753 ** Get the list of tracepoints.
1754 ** Share an <<event,event rule>> so that the user space tracing library
1755 can enable or disable tracepoints. Amongst the possible conditions
1756 of an event rule is a filter expression which `liblttng-ust` evalutes
1757 when an event occurs.
1758 ** Share <<channel,channel>> attributes and ring buffer locations.
1761 The session daemon and the user space tracing library use a Unix
1762 domain socket for their communication.
1764 * The <<lttng-ust-agents,user space tracing agents>>.
1766 Any instance of a user space tracing agent first registers to
1767 a session daemon. Then, the session daemon can send requests to
1768 this instance, such as:
1771 ** Get the list of loggers.
1772 ** Enable or disable a specific logger.
1775 The session daemon and the user space tracing agent use a TCP connection
1776 for their communication.
1778 * The <<lttng-modules,LTTng kernel tracer>>.
1779 * The <<lttng-consumerd,consumer daemon>>.
1781 The session daemon sends requests to the consumer daemon to instruct
1782 it where to send the trace data streams, amongst other information.
1784 * The <<lttng-relayd,relay daemon>>.
1786 The session daemon receives commands from the
1787 <<liblttng-ctl-lttng,tracing control library>>.
1789 The root session daemon loads the appropriate
1790 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
1791 a <<lttng-consumerd,consumer daemon>> as soon as you create
1792 an <<event,event rule>>.
1794 The session daemon does not send and receive trace data: this is the
1795 role of the <<lttng-consumerd,consumer daemon>> and
1796 <<lttng-relayd,relay daemon>>. It does, however, generate the
1797 http://diamon.org/ctf/[CTF] metadata stream.
1799 Each Unix user can have its own session daemon instance. The
1800 tracing sessions which different session daemons manage are completely
1803 The root user's session daemon is the only one which is
1804 allowed to control the LTTng kernel tracer, and its spawned consumer
1805 daemon is the only one which is allowed to consume trace data from the
1806 LTTng kernel tracer. Note, however, that any Unix user which is a member
1807 of the <<tracing-group,tracing group>> is allowed
1808 to create <<channel,channels>> in the
1809 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
1812 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
1813 session daemon when using its `create` command if none is currently
1814 running. You can also start the session daemon manually.
1821 .The consumer daemon.
1822 image::plumbing-consumerd.png[]
1824 The _consumer daemon_, cmd:lttng-consumerd, is a daemon which shares
1825 ring buffers with user applications or with the LTTng kernel modules to
1826 collect trace data and send it to some location (on disk or to a
1827 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
1828 is part of LTTng-tools.
1830 You do not start a consumer daemon manually: a consumer daemon is always
1831 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
1832 <<event,event rule>>, that is, before you start tracing. When you kill
1833 its owner session daemon, the consumer daemon also exits because it is
1834 the session daemon's child process. Command-line options of
1835 man:lttng-sessiond(8) target the consumer daemon process.
1837 There are up to two running consumer daemons per Unix user, whereas only
1838 one session daemon can run per user. This is because each process can be
1839 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
1840 and 64-bit processes, it is more efficient to have separate
1841 corresponding 32-bit and 64-bit consumer daemons. The root user is an
1842 exception: it can have up to _three_ running consumer daemons: 32-bit
1843 and 64-bit instances for its user applications, and one more
1844 reserved for collecting kernel trace data.
1852 image::plumbing-relayd.png[]
1854 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
1855 between remote session and consumer daemons, local trace files, and a
1856 remote live trace viewer. The relay daemon is part of LTTng-tools.
1858 The main purpose of the relay daemon is to implement a receiver of
1859 <<sending-trace-data-over-the-network,trace data over the network>>.
1860 This is useful when the target system does not have much file system
1861 space to record trace files locally.
1863 The relay daemon is also a server to which a
1864 <<lttng-live,live trace viewer>> can
1865 connect. The live trace viewer sends requests to the relay daemon to
1866 receive trace data as the target system emits events. The
1867 communication protocol is named _LTTng live_; it is used over TCP
1870 Note that you can start the relay daemon on the target system directly.
1871 This is the setup of choice when the use case is to view events as
1872 the target system emits them without the need of a remote system.
1876 == [[using-lttng]]Instrumentation
1878 There are many examples of tracing and monitoring in our everyday life:
1880 * You have access to real-time and historical weather reports and
1881 forecasts thanks to weather stations installed around the country.
1882 * You know your heart is safe thanks to an electrocardiogram.
1883 * You make sure not to drive your car too fast and to have enough fuel
1884 to reach your destination thanks to gauges visible on your dashboard.
1886 All the previous examples have something in common: they rely on
1887 **instruments**. Without the electrodes attached to the surface of your
1888 body's skin, cardiac monitoring is futile.
1890 LTTng, as a tracer, is no different from those real life examples. If
1891 you're about to trace a software system or, in other words, record its
1892 history of execution, you better have **instrumentation points** in the
1893 subject you're tracing, that is, the actual software.
1895 Various ways were developed to instrument a piece of software for LTTng
1896 tracing. The most straightforward one is to manually place
1897 instrumentation points, called _tracepoints_, in the software's source
1898 code. It is also possible to add instrumentation points dynamically in
1899 the Linux kernel <<domain,tracing domain>>.
1901 If you're only interested in tracing the Linux kernel, your
1902 instrumentation needs are probably already covered by LTTng's built-in
1903 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
1904 user application which is already instrumented for LTTng tracing.
1905 In such cases, you can skip this whole section and read the topics of
1906 the <<controlling-tracing,Tracing control>> section.
1908 Many methods are available to instrument a piece of software for LTTng
1911 * <<c-application,User space instrumentation for C and $$C++$$
1913 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
1914 * <<java-application,User space Java agent>>.
1915 * <<python-application,User space Python agent>>.
1916 * <<proc-lttng-logger-abi,LTTng logger>>.
1917 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
1921 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
1923 The procedure to instrument a C or $$C++$$ user application with
1924 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
1926 . <<tracepoint-provider,Create the source files of a tracepoint provider
1928 . <<probing-the-application-source-code,Add tracepoints to
1929 the application's source code>>.
1930 . <<building-tracepoint-providers-and-user-application,Build and link
1931 a tracepoint provider package and the user application>>.
1933 If you need quick, man:printf(3)-like instrumentation, you can skip
1934 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
1937 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
1938 instrument a user application with `liblttng-ust`.
1941 [[tracepoint-provider]]
1942 ==== Create the source files of a tracepoint provider package
1944 A _tracepoint provider_ is a set of compiled functions which provide
1945 **tracepoints** to an application, the type of instrumentation point
1946 supported by LTTng-UST. Those functions can emit events with
1947 user-defined fields and serialize those events as event records to one
1948 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
1949 macro, which you <<probing-the-application-source-code,insert in a user
1950 application's source code>>, calls those functions.
1952 A _tracepoint provider package_ is an object file (`.o`) or a shared
1953 library (`.so`) which contains one or more tracepoint providers.
1954 Its source files are:
1956 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
1957 * A <<tpp-source,tracepoint provider package source>> (`.c`).
1959 A tracepoint provider package is dynamically linked with `liblttng-ust`,
1960 the LTTng user space tracer, at run time.
1963 .User application linked with `liblttng-ust` and containing a tracepoint provider.
1964 image::ust-app.png[]
1966 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
1967 skip creating and using a tracepoint provider and use
1968 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
1972 ===== Create a tracepoint provider header file template
1974 A _tracepoint provider header file_ contains the tracepoint
1975 definitions of a tracepoint provider.
1977 To create a tracepoint provider header file:
1979 . Start from this template:
1983 .Tracepoint provider header file template (`.h` file extension).
1985 #undef TRACEPOINT_PROVIDER
1986 #define TRACEPOINT_PROVIDER provider_name
1988 #undef TRACEPOINT_INCLUDE
1989 #define TRACEPOINT_INCLUDE "./tp.h"
1991 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
1994 #include <lttng/tracepoint.h>
1997 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
1998 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2003 #include <lttng/tracepoint-event.h>
2009 * `provider_name` with the name of your tracepoint provider.
2010 * `"tp.h"` with the name of your tracepoint provider header file.
2012 . Below the `#include <lttng/tracepoint.h>` line, put your
2013 <<defining-tracepoints,tracepoint definitions>>.
2015 Your tracepoint provider name must be unique amongst all the possible
2016 tracepoint provider names used on the same target system. We
2017 suggest to include the name of your project or company in the name,
2018 for example, `org_lttng_my_project_tpp`.
2020 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2021 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2022 write are the <<defining-tracepoints,tracepoint definitions>>.
2025 [[defining-tracepoints]]
2026 ===== Create a tracepoint definition
2028 A _tracepoint definition_ defines, for a given tracepoint:
2030 * Its **input arguments**. They are the macro parameters that the
2031 `tracepoint()` macro accepts for this particular tracepoint
2032 in the user application's source code.
2033 * Its **output event fields**. They are the sources of event fields
2034 that form the payload of any event that the execution of the
2035 `tracepoint()` macro emits for this particular tracepoint.
2037 You can create a tracepoint definition by using the
2038 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2040 <<tpp-header,tracepoint provider header file template>>.
2042 The syntax of the `TRACEPOINT_EVENT()` macro is:
2045 .`TRACEPOINT_EVENT()` macro syntax.
2048 /* Tracepoint provider name */
2051 /* Tracepoint name */
2054 /* Input arguments */
2059 /* Output event fields */
2068 * `provider_name` with your tracepoint provider name.
2069 * `tracepoint_name` with your tracepoint name.
2070 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2071 * `fields` with the <<tpp-def-output-fields,output event field>>
2074 This tracepoint emits events named `provider_name:tracepoint_name`.
2077 .Event name's length limitation
2079 The concatenation of the tracepoint provider name and the
2080 tracepoint name must not exceed **254{nbsp}characters**. If it does, the
2081 instrumented application compiles and runs, but LTTng throws multiple
2082 warnings and you could experience serious issues.
2085 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2088 .`TP_ARGS()` macro syntax.
2097 * `type` with the C type of the argument.
2098 * `arg_name` with the argument name.
2100 You can repeat `type` and `arg_name` up to 10{nbsp}times to have
2101 more than one argument.
2103 .`TP_ARGS()` usage with three arguments.
2115 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2116 tracepoint definition with no input arguments.
2118 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2119 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2120 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2121 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2124 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2125 C expression that the tracer evalutes at the `tracepoint()` macro site
2126 in the application's source code. This expression provides a field's
2127 source of data. The argument expression can include input argument names
2128 listed in the `TP_ARGS()` macro.
2130 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2131 must be unique within a given tracepoint definition.
2133 Here's a complete tracepoint definition example:
2135 .Tracepoint definition.
2137 The following tracepoint definition defines a tracepoint which takes
2138 three input arguments and has four output event fields.
2142 #include "my-custom-structure.h"
2148 const struct my_custom_structure*, my_custom_structure,
2153 ctf_string(query_field, query)
2154 ctf_float(double, ratio_field, ratio)
2155 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2156 ctf_integer(int, send_size, my_custom_structure->send_size)
2161 You can refer to this tracepoint definition with the `tracepoint()`
2162 macro in your application's source code like this:
2166 tracepoint(my_provider, my_tracepoint,
2167 my_structure, some_ratio, the_query);
2171 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2172 if they satisfy an enabled <<event,event rule>>.
2175 [[using-tracepoint-classes]]
2176 ===== Use a tracepoint class
2178 A _tracepoint class_ is a class of tracepoints which share the same
2179 output event field definitions. A _tracepoint instance_ is one
2180 instance of such a defined tracepoint class, with its own tracepoint
2183 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2184 shorthand which defines both a tracepoint class and a tracepoint
2185 instance at the same time.
2187 When you build a tracepoint provider package, the C or $$C++$$ compiler
2188 creates one serialization function for each **tracepoint class**. A
2189 serialization function is responsible for serializing the event fields
2190 of a tracepoint to a sub-buffer when tracing.
2192 For various performance reasons, when your situation requires multiple
2193 tracepoint definitions with different names, but with the same event
2194 fields, we recommend that you manually create a tracepoint class
2195 and instantiate as many tracepoint instances as needed. One positive
2196 effect of such a design, amongst other advantages, is that all
2197 tracepoint instances of the same tracepoint class reuse the same
2198 serialization function, thus reducing
2199 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2201 .Use a tracepoint class and tracepoint instances.
2203 Consider the following three tracepoint definitions:
2215 ctf_integer(int, userid, userid)
2216 ctf_integer(size_t, len, len)
2228 ctf_integer(int, userid, userid)
2229 ctf_integer(size_t, len, len)
2241 ctf_integer(int, userid, userid)
2242 ctf_integer(size_t, len, len)
2247 In this case, we create three tracepoint classes, with one implicit
2248 tracepoint instance for each of them: `get_account`, `get_settings`, and
2249 `get_transaction`. However, they all share the same event field names
2250 and types. Hence three identical, yet independent serialization
2251 functions are created when you build the tracepoint provider package.
2253 A better design choice is to define a single tracepoint class and three
2254 tracepoint instances:
2258 /* The tracepoint class */
2259 TRACEPOINT_EVENT_CLASS(
2260 /* Tracepoint provider name */
2263 /* Tracepoint class name */
2266 /* Input arguments */
2272 /* Output event fields */
2274 ctf_integer(int, userid, userid)
2275 ctf_integer(size_t, len, len)
2279 /* The tracepoint instances */
2280 TRACEPOINT_EVENT_INSTANCE(
2281 /* Tracepoint provider name */
2284 /* Tracepoint class name */
2287 /* Tracepoint name */
2290 /* Input arguments */
2296 TRACEPOINT_EVENT_INSTANCE(
2305 TRACEPOINT_EVENT_INSTANCE(
2318 [[assigning-log-levels]]
2319 ===== Assign a log level to a tracepoint definition
2321 You can assign an optional _log level_ to a
2322 <<defining-tracepoints,tracepoint definition>>.
2324 Assigning different levels of severity to tracepoint definitions can
2325 be useful: when you <<enabling-disabling-events,create an event rule>>,
2326 you can target tracepoints having a log level as severe as a specific
2329 The concept of LTTng-UST log levels is similar to the levels found
2330 in typical logging frameworks:
2332 * In a logging framework, the log level is given by the function
2333 or method name you use at the log statement site: `debug()`,
2334 `info()`, `warn()`, `error()`, and so on.
2335 * In LTTng-UST, you statically assign the log level to a tracepoint
2336 definition; any `tracepoint()` macro invocation which refers to
2337 this definition has this log level.
2339 You can assign a log level to a tracepoint definition with the
2340 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2341 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2342 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2345 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2348 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2350 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2355 * `provider_name` with the tracepoint provider name.
2356 * `tracepoint_name` with the tracepoint name.
2357 * `log_level` with the log level to assign to the tracepoint
2358 definition named `tracepoint_name` in the `provider_name`
2359 tracepoint provider.
2361 See man:lttng-ust(3) for a list of available log level names.
2363 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2367 /* Tracepoint definition */
2376 ctf_integer(int, userid, userid)
2377 ctf_integer(size_t, len, len)
2381 /* Log level assignment */
2382 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2388 ===== Create a tracepoint provider package source file
2390 A _tracepoint provider package source file_ is a C source file which
2391 includes a <<tpp-header,tracepoint provider header file>> to expand its
2392 macros into event serialization and other functions.
2394 You can always use the following tracepoint provider package source
2398 .Tracepoint provider package source file template.
2400 #define TRACEPOINT_CREATE_PROBES
2405 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2406 header file>> name. You may also include more than one tracepoint
2407 provider header file here to create a tracepoint provider package
2408 holding more than one tracepoint providers.
2411 [[probing-the-application-source-code]]
2412 ==== Add tracepoints to an application's source code
2414 Once you <<tpp-header,create a tracepoint provider header file>>, you
2415 can use the `tracepoint()` macro in your application's
2416 source code to insert the tracepoints that this header
2417 <<defining-tracepoints,defines>>.
2419 The `tracepoint()` macro takes at least two parameters: the tracepoint
2420 provider name and the tracepoint name. The corresponding tracepoint
2421 definition defines the other parameters.
2423 .`tracepoint()` usage.
2425 The following <<defining-tracepoints,tracepoint definition>> defines a
2426 tracepoint which takes two input arguments and has two output event
2430 .Tracepoint provider header file.
2432 #include "my-custom-structure.h"
2439 const char*, cmd_name
2442 ctf_string(cmd_name, cmd_name)
2443 ctf_integer(int, number_of_args, argc)
2448 You can refer to this tracepoint definition with the `tracepoint()`
2449 macro in your application's source code like this:
2452 .Application's source file.
2456 int main(int argc, char* argv[])
2458 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2464 Note how the application's source code includes
2465 the tracepoint provider header file containing the tracepoint
2466 definitions to use, path:{tp.h}.
2469 .`tracepoint()` usage with a complex tracepoint definition.
2471 Consider this complex tracepoint definition, where multiple event
2472 fields refer to the same input arguments in their argument expression
2476 .Tracepoint provider header file.
2478 /* For `struct stat` */
2479 #include <sys/types.h>
2480 #include <sys/stat.h>
2492 ctf_integer(int, my_constant_field, 23 + 17)
2493 ctf_integer(int, my_int_arg_field, my_int_arg)
2494 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2495 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2496 my_str_arg[2] + my_str_arg[3])
2497 ctf_string(my_str_arg_field, my_str_arg)
2498 ctf_integer_hex(off_t, size_field, st->st_size)
2499 ctf_float(double, size_dbl_field, (double) st->st_size)
2500 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2501 size_t, strlen(my_str_arg) / 2)
2506 You can refer to this tracepoint definition with the `tracepoint()`
2507 macro in your application's source code like this:
2510 .Application's source file.
2512 #define TRACEPOINT_DEFINE
2519 stat("/etc/fstab", &s);
2520 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2526 If you look at the event record that LTTng writes when tracing this
2527 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2528 it should look like this:
2530 .Event record fields
2532 |Field's name |Field's value
2533 |`my_constant_field` |40
2534 |`my_int_arg_field` |23
2535 |`my_int_arg_field2` |529
2537 |`my_str_arg_field` |`Hello, World!`
2538 |`size_field` |0x12d
2539 |`size_dbl_field` |301.0
2540 |`half_my_str_arg_field` |`Hello,`
2544 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2545 compute--they use the call stack, for example. To avoid this
2546 computation when the tracepoint is disabled, you can use the
2547 `tracepoint_enabled()` and `do_tracepoint()` macros.
2549 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2553 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2555 tracepoint_enabled(provider_name, tracepoint_name)
2556 do_tracepoint(provider_name, tracepoint_name, ...)
2561 * `provider_name` with the tracepoint provider name.
2562 * `tracepoint_name` with the tracepoint name.
2564 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2565 `tracepoint_name` from the provider named `provider_name` is enabled
2568 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2569 if the tracepoint is enabled. Using `tracepoint()` with
2570 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2571 the `tracepoint_enabled()` check, thus a race condition is
2572 possible in this situation:
2575 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2577 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2578 stuff = prepare_stuff();
2581 tracepoint(my_provider, my_tracepoint, stuff);
2584 If the tracepoint is enabled after the condition, then `stuff` is not
2585 prepared: the emitted event will either contain wrong data, or the whole
2586 application could crash (segmentation fault, for example).
2588 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2589 `STAP_PROBEV()` call. If you need it, you must emit
2593 [[building-tracepoint-providers-and-user-application]]
2594 ==== Build and link a tracepoint provider package and an application
2596 Once you have one or more <<tpp-header,tracepoint provider header
2597 files>> and a <<tpp-source,tracepoint provider package source file>>,
2598 you can create the tracepoint provider package by compiling its source
2599 file. From here, multiple build and run scenarios are possible. The
2600 following table shows common application and library configurations
2601 along with the required command lines to achieve them.
2603 In the following diagrams, we use the following file names:
2606 Executable application.
2609 Application's object file.
2612 Tracepoint provider package object file.
2615 Tracepoint provider package archive file.
2618 Tracepoint provider package shared object file.
2621 User library object file.
2624 User library shared object file.
2626 We use the following symbols in the diagrams of table below:
2629 .Symbols used in the build scenario diagrams.
2630 image::ust-sit-symbols.png[]
2632 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2633 variable in the following instructions.
2635 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2636 .Common tracepoint provider package scenarios.
2638 |Scenario |Instructions
2641 The instrumented application is statically linked with
2642 the tracepoint provider package object.
2644 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2647 include::../common/ust-sit-step-tp-o.txt[]
2649 To build the instrumented application:
2651 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2656 #define TRACEPOINT_DEFINE
2660 . Compile the application source file:
2669 . Build the application:
2674 $ gcc -o app app.o tpp.o -llttng-ust -ldl
2678 To run the instrumented application:
2680 * Start the application:
2690 The instrumented application is statically linked with the
2691 tracepoint provider package archive file.
2693 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2696 To create the tracepoint provider package archive file:
2698 . Compile the <<tpp-source,tracepoint provider package source file>>:
2707 . Create the tracepoint provider package archive file:
2712 $ ar rcs tpp.a tpp.o
2716 To build the instrumented application:
2718 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2723 #define TRACEPOINT_DEFINE
2727 . Compile the application source file:
2736 . Build the application:
2741 $ gcc -o app app.o tpp.a -llttng-ust -ldl
2745 To run the instrumented application:
2747 * Start the application:
2757 The instrumented application is linked with the tracepoint provider
2758 package shared object.
2760 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
2763 include::../common/ust-sit-step-tp-so.txt[]
2765 To build the instrumented application:
2767 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2772 #define TRACEPOINT_DEFINE
2776 . Compile the application source file:
2785 . Build the application:
2790 $ gcc -o app app.o -ldl -L. -ltpp
2794 To run the instrumented application:
2796 * Start the application:
2806 The tracepoint provider package shared object is preloaded before the
2807 instrumented application starts.
2809 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
2812 include::../common/ust-sit-step-tp-so.txt[]
2814 To build the instrumented application:
2816 . In path:{app.c}, before including path:{tpp.h}, add the
2822 #define TRACEPOINT_DEFINE
2823 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2827 . Compile the application source file:
2836 . Build the application:
2841 $ gcc -o app app.o -ldl
2845 To run the instrumented application with tracing support:
2847 * Preload the tracepoint provider package shared object and
2848 start the application:
2853 $ LD_PRELOAD=./libtpp.so ./app
2857 To run the instrumented application without tracing support:
2859 * Start the application:
2869 The instrumented application dynamically loads the tracepoint provider
2870 package shared object.
2872 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
2875 include::../common/ust-sit-step-tp-so.txt[]
2877 To build the instrumented application:
2879 . In path:{app.c}, before including path:{tpp.h}, add the
2885 #define TRACEPOINT_DEFINE
2886 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
2890 . Compile the application source file:
2899 . Build the application:
2904 $ gcc -o app app.o -ldl
2908 To run the instrumented application:
2910 * Start the application:
2920 The application is linked with the instrumented user library.
2922 The instrumented user library is statically linked with the tracepoint
2923 provider package object file.
2925 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
2928 include::../common/ust-sit-step-tp-o-fpic.txt[]
2930 To build the instrumented user library:
2932 . In path:{emon.c}, before including path:{tpp.h}, add the
2938 #define TRACEPOINT_DEFINE
2942 . Compile the user library source file:
2947 $ gcc -I. -fpic -c emon.c
2951 . Build the user library shared object:
2956 $ gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
2960 To build the application:
2962 . Compile the application source file:
2971 . Build the application:
2976 $ gcc -o app app.o -L. -lemon
2980 To run the application:
2982 * Start the application:
2992 The application is linked with the instrumented user library.
2994 The instrumented user library is linked with the tracepoint provider
2995 package shared object.
2997 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3000 include::../common/ust-sit-step-tp-so.txt[]
3002 To build the instrumented user library:
3004 . In path:{emon.c}, before including path:{tpp.h}, add the
3010 #define TRACEPOINT_DEFINE
3014 . Compile the user library source file:
3019 $ gcc -I. -fpic -c emon.c
3023 . Build the user library shared object:
3028 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3032 To build the application:
3034 . Compile the application source file:
3043 . Build the application:
3048 $ gcc -o app app.o -L. -lemon
3052 To run the application:
3054 * Start the application:
3064 The tracepoint provider package shared object is preloaded before the
3067 The application is linked with the instrumented user library.
3069 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3072 include::../common/ust-sit-step-tp-so.txt[]
3074 To build the instrumented user library:
3076 . In path:{emon.c}, before including path:{tpp.h}, add the
3082 #define TRACEPOINT_DEFINE
3083 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3087 . Compile the user library source file:
3092 $ gcc -I. -fpic -c emon.c
3096 . Build the user library shared object:
3101 $ gcc -shared -o libemon.so emon.o -ldl
3105 To build the application:
3107 . Compile the application source file:
3116 . Build the application:
3121 $ gcc -o app app.o -L. -lemon
3125 To run the application with tracing support:
3127 * Preload the tracepoint provider package shared object and
3128 start the application:
3133 $ LD_PRELOAD=./libtpp.so ./app
3137 To run the application without tracing support:
3139 * Start the application:
3149 The application is linked with the instrumented user library.
3151 The instrumented user library dynamically loads the tracepoint provider
3152 package shared object.
3154 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3157 include::../common/ust-sit-step-tp-so.txt[]
3159 To build the instrumented user library:
3161 . In path:{emon.c}, before including path:{tpp.h}, add the
3167 #define TRACEPOINT_DEFINE
3168 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3172 . Compile the user library source file:
3177 $ gcc -I. -fpic -c emon.c
3181 . Build the user library shared object:
3186 $ gcc -shared -o libemon.so emon.o -ldl
3190 To build the application:
3192 . Compile the application source file:
3201 . Build the application:
3206 $ gcc -o app app.o -L. -lemon
3210 To run the application:
3212 * Start the application:
3222 The application dynamically loads the instrumented user library.
3224 The instrumented user library is linked with the tracepoint provider
3225 package shared object.
3227 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3230 include::../common/ust-sit-step-tp-so.txt[]
3232 To build the instrumented user library:
3234 . In path:{emon.c}, before including path:{tpp.h}, add the
3240 #define TRACEPOINT_DEFINE
3244 . Compile the user library source file:
3249 $ gcc -I. -fpic -c emon.c
3253 . Build the user library shared object:
3258 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3262 To build the application:
3264 . Compile the application source file:
3273 . Build the application:
3278 $ gcc -o app app.o -ldl -L. -lemon
3282 To run the application:
3284 * Start the application:
3294 The application dynamically loads the instrumented user library.
3296 The instrumented user library dynamically loads the tracepoint provider
3297 package shared object.
3299 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3302 include::../common/ust-sit-step-tp-so.txt[]
3304 To build the instrumented user library:
3306 . In path:{emon.c}, before including path:{tpp.h}, add the
3312 #define TRACEPOINT_DEFINE
3313 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3317 . Compile the user library source file:
3322 $ gcc -I. -fpic -c emon.c
3326 . Build the user library shared object:
3331 $ gcc -shared -o libemon.so emon.o -ldl
3335 To build the application:
3337 . Compile the application source file:
3346 . Build the application:
3351 $ gcc -o app app.o -ldl -L. -lemon
3355 To run the application:
3357 * Start the application:
3367 The tracepoint provider package shared object is preloaded before the
3370 The application dynamically loads the instrumented user library.
3372 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3375 include::../common/ust-sit-step-tp-so.txt[]
3377 To build the instrumented user library:
3379 . In path:{emon.c}, before including path:{tpp.h}, add the
3385 #define TRACEPOINT_DEFINE
3386 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3390 . Compile the user library source file:
3395 $ gcc -I. -fpic -c emon.c
3399 . Build the user library shared object:
3404 $ gcc -shared -o libemon.so emon.o -ldl
3408 To build the application:
3410 . Compile the application source file:
3419 . Build the application:
3424 $ gcc -o app app.o -L. -lemon
3428 To run the application with tracing support:
3430 * Preload the tracepoint provider package shared object and
3431 start the application:
3436 $ LD_PRELOAD=./libtpp.so ./app
3440 To run the application without tracing support:
3442 * Start the application:
3452 The application is statically linked with the tracepoint provider
3453 package object file.
3455 The application is linked with the instrumented user library.
3457 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3460 include::../common/ust-sit-step-tp-o.txt[]
3462 To build the instrumented user library:
3464 . In path:{emon.c}, before including path:{tpp.h}, add the
3470 #define TRACEPOINT_DEFINE
3474 . Compile the user library source file:
3479 $ gcc -I. -fpic -c emon.c
3483 . Build the user library shared object:
3488 $ gcc -shared -o libemon.so emon.o
3492 To build the application:
3494 . Compile the application source file:
3503 . Build the application:
3508 $ gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3512 To run the instrumented application:
3514 * Start the application:
3524 The application is statically linked with the tracepoint provider
3525 package object file.
3527 The application dynamically loads the instrumented user library.
3529 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3532 include::../common/ust-sit-step-tp-o.txt[]
3534 To build the application:
3536 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3541 #define TRACEPOINT_DEFINE
3545 . Compile the application source file:
3554 . Build the application:
3559 $ gcc -Wl,--export-dynamic -o app app.o tpp.o \
3564 The `--export-dynamic` option passed to the linker is necessary for the
3565 dynamically loaded library to ``see'' the tracepoint symbols defined in
3568 To build the instrumented user library:
3570 . Compile the user library source file:
3575 $ gcc -I. -fpic -c emon.c
3579 . Build the user library shared object:
3584 $ gcc -shared -o libemon.so emon.o
3588 To run the application:
3590 * Start the application:
3601 [[using-lttng-ust-with-daemons]]
3602 ===== Use noch:{LTTng-UST} with daemons
3604 If your instrumented application calls man:fork(2), man:clone(2),
3605 or BSD's man:rfork(2), without a following man:exec(3)-family
3606 system call, you must preload the path:{liblttng-ust-fork.so} shared
3607 object when you start the application.
3611 $ LD_PRELOAD=liblttng-ust-fork.so ./my-app
3614 If your tracepoint provider package is
3615 a shared library which you also preload, you must put both
3616 shared objects in env:LD_PRELOAD:
3620 $ LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3626 ===== Use noch:{LTTng-UST} with applications which close file descriptors that don't belong to them
3628 If your instrumented application closes one or more file descriptors
3629 which it did not open itself, you must preload the
3630 path:{liblttng-ust-fd.so} shared object when you start the application:
3634 $ LD_PRELOAD=liblttng-ust-fd.so ./my-app
3637 Typical use cases include closing all the file descriptors after
3638 man:fork(2) or man:rfork(2) and buggy applications doing
3642 [[lttng-ust-pkg-config]]
3643 ===== Use noch:{pkg-config}
3645 On some distributions, LTTng-UST ships with a
3646 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3647 metadata file. If this is your case, then you can use cmd:pkg-config to
3648 build an application on the command line:
3652 $ gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3656 [[instrumenting-32-bit-app-on-64-bit-system]]
3657 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3659 In order to trace a 32-bit application running on a 64-bit system,
3660 LTTng must use a dedicated 32-bit
3661 <<lttng-consumerd,consumer daemon>>.
3663 The following steps show how to build and install a 32-bit consumer
3664 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3665 build and install the 32-bit LTTng-UST libraries, and how to build and
3666 link an instrumented 32-bit application in that context.
3668 To build a 32-bit instrumented application for a 64-bit target system,
3669 assuming you have a fresh target system with no installed Userspace RCU
3672 . Download, build, and install a 32-bit version of Userspace RCU:
3677 $ cd $(mktemp -d) &&
3678 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3679 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3680 cd userspace-rcu-0.9.* &&
3681 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3683 sudo make install &&
3688 . Using your distribution's package manager, or from source, install
3689 the following 32-bit versions of the following dependencies of
3690 LTTng-tools and LTTng-UST:
3693 * https://sourceforge.net/projects/libuuid/[libuuid]
3694 * http://directory.fsf.org/wiki/Popt[popt]
3695 * http://www.xmlsoft.org/[libxml2]
3698 . Download, build, and install a 32-bit version of the latest
3699 LTTng-UST{nbsp}{revision}:
3704 $ cd $(mktemp -d) &&
3705 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.11.tar.bz2 &&
3706 tar -xf lttng-ust-latest-2.11.tar.bz2 &&
3707 cd lttng-ust-2.11.* &&
3708 ./configure --libdir=/usr/local/lib32 \
3709 CFLAGS=-m32 CXXFLAGS=-m32 \
3710 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
3712 sudo make install &&
3719 Depending on your distribution,
3720 32-bit libraries could be installed at a different location than
3721 `/usr/lib32`. For example, Debian is known to install
3722 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
3724 In this case, make sure to set `LDFLAGS` to all the
3725 relevant 32-bit library paths, for example:
3729 $ LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
3733 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
3734 the 32-bit consumer daemon:
3739 $ cd $(mktemp -d) &&
3740 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.11.tar.bz2 &&
3741 tar -xf lttng-tools-latest-2.11.tar.bz2 &&
3742 cd lttng-tools-2.11.* &&
3743 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
3744 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' \
3745 --disable-bin-lttng --disable-bin-lttng-crash \
3746 --disable-bin-lttng-relayd --disable-bin-lttng-sessiond &&
3748 cd src/bin/lttng-consumerd &&
3749 sudo make install &&
3754 . From your distribution or from source,
3755 <<installing-lttng,install>> the 64-bit versions of
3756 LTTng-UST and Userspace RCU.
3757 . Download, build, and install the 64-bit version of the
3758 latest LTTng-tools{nbsp}{revision}:
3763 $ cd $(mktemp -d) &&
3764 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.11.tar.bz2 &&
3765 tar -xf lttng-tools-latest-2.11.tar.bz2 &&
3766 cd lttng-tools-2.11.* &&
3767 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
3768 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
3770 sudo make install &&
3775 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
3776 when linking your 32-bit application:
3779 -m32 -L/usr/lib32 -L/usr/local/lib32 \
3780 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
3783 For example, let's rebuild the quick start example in
3784 <<tracing-your-own-user-application,Trace a user application>> as an
3785 instrumented 32-bit application:
3790 $ gcc -m32 -c -I. hello-tp.c
3791 $ gcc -m32 -c hello.c
3792 $ gcc -m32 -o hello hello.o hello-tp.o \
3793 -L/usr/lib32 -L/usr/local/lib32 \
3794 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
3799 No special action is required to execute the 32-bit application and
3800 to trace it: use the command-line man:lttng(1) tool as usual.
3807 man:tracef(3) is a small LTTng-UST API designed for quick,
3808 man:printf(3)-like instrumentation without the burden of
3809 <<tracepoint-provider,creating>> and
3810 <<building-tracepoint-providers-and-user-application,building>>
3811 a tracepoint provider package.
3813 To use `tracef()` in your application:
3815 . In the C or C++ source files where you need to use `tracef()`,
3816 include `<lttng/tracef.h>`:
3821 #include <lttng/tracef.h>
3825 . In the application's source code, use `tracef()` like you would use
3833 tracef("my message: %d (%s)", my_integer, my_string);
3839 . Link your application with `liblttng-ust`:
3844 $ gcc -o app app.c -llttng-ust
3848 To trace the events that `tracef()` calls emit:
3850 * <<enabling-disabling-events,Create an event rule>> which matches the
3851 `lttng_ust_tracef:*` event name:
3856 $ lttng enable-event --userspace 'lttng_ust_tracef:*'
3861 .Limitations of `tracef()`
3863 The `tracef()` utility function was developed to make user space tracing
3864 super simple, albeit with notable disadvantages compared to
3865 <<defining-tracepoints,user-defined tracepoints>>:
3867 * All the emitted events have the same tracepoint provider and
3868 tracepoint names, respectively `lttng_ust_tracef` and `event`.
3869 * There is no static type checking.
3870 * The only event record field you actually get, named `msg`, is a string
3871 potentially containing the values you passed to `tracef()`
3872 using your own format string. This also means that you cannot filter
3873 events with a custom expression at run time because there are no
3875 * Since `tracef()` uses the C standard library's man:vasprintf(3)
3876 function behind the scenes to format the strings at run time, its
3877 expected performance is lower than with user-defined tracepoints,
3878 which do not require a conversion to a string.
3880 Taking this into consideration, `tracef()` is useful for some quick
3881 prototyping and debugging, but you should not consider it for any
3882 permanent and serious applicative instrumentation.
3888 ==== Use `tracelog()`
3890 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
3891 the difference that it accepts an additional log level parameter.
3893 The goal of `tracelog()` is to ease the migration from logging to
3896 To use `tracelog()` in your application:
3898 . In the C or C++ source files where you need to use `tracelog()`,
3899 include `<lttng/tracelog.h>`:
3904 #include <lttng/tracelog.h>
3908 . In the application's source code, use `tracelog()` like you would use
3909 man:printf(3), except for the first parameter which is the log
3917 tracelog(TRACE_WARNING, "my message: %d (%s)",
3918 my_integer, my_string);
3924 See man:lttng-ust(3) for a list of available log level names.
3926 . Link your application with `liblttng-ust`:
3931 $ gcc -o app app.c -llttng-ust
3935 To trace the events that `tracelog()` calls emit with a log level
3936 _as severe as_ a specific log level:
3938 * <<enabling-disabling-events,Create an event rule>> which matches the
3939 `lttng_ust_tracelog:*` event name and a minimum level
3945 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
3946 --loglevel=TRACE_WARNING
3950 To trace the events that `tracelog()` calls emit with a
3951 _specific log level_:
3953 * Create an event rule which matches the `lttng_ust_tracelog:*`
3954 event name and a specific log level:
3959 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
3960 --loglevel-only=TRACE_INFO
3965 [[prebuilt-ust-helpers]]
3966 === Prebuilt user space tracing helpers
3968 The LTTng-UST package provides a few helpers in the form or preloadable
3969 shared objects which automatically instrument system functions and
3972 The helper shared objects are normally found in dir:{/usr/lib}. If you
3973 built LTTng-UST <<building-from-source,from source>>, they are probably
3974 located in dir:{/usr/local/lib}.
3976 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
3979 path:{liblttng-ust-libc-wrapper.so}::
3980 path:{liblttng-ust-pthread-wrapper.so}::
3981 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
3982 memory and POSIX threads function tracing>>.
3984 path:{liblttng-ust-cyg-profile.so}::
3985 path:{liblttng-ust-cyg-profile-fast.so}::
3986 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
3988 path:{liblttng-ust-dl.so}::
3989 <<liblttng-ust-dl,Dynamic linker tracing>>.
3991 To use a user space tracing helper with any user application:
3993 * Preload the helper shared object when you start the application:
3998 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4002 You can preload more than one helper:
4007 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4013 [[liblttng-ust-libc-pthread-wrapper]]
4014 ==== Instrument C standard library memory and POSIX threads functions
4016 The path:{liblttng-ust-libc-wrapper.so} and
4017 path:{liblttng-ust-pthread-wrapper.so} helpers
4018 add instrumentation to some C standard library and POSIX
4022 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4024 |TP provider name |TP name |Instrumented function
4026 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4027 |`calloc` |man:calloc(3)
4028 |`realloc` |man:realloc(3)
4029 |`free` |man:free(3)
4030 |`memalign` |man:memalign(3)
4031 |`posix_memalign` |man:posix_memalign(3)
4035 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4037 |TP provider name |TP name |Instrumented function
4039 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4040 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4041 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4042 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4045 When you preload the shared object, it replaces the functions listed
4046 in the previous tables by wrappers which contain tracepoints and call
4047 the replaced functions.
4050 [[liblttng-ust-cyg-profile]]
4051 ==== Instrument function entry and exit
4053 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4054 to the entry and exit points of functions.
4056 man:gcc(1) and man:clang(1) have an option named
4057 https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html[`-finstrument-functions`]
4058 which generates instrumentation calls for entry and exit to functions.
4059 The LTTng-UST function tracing helpers,
4060 path:{liblttng-ust-cyg-profile.so} and
4061 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4062 to add tracepoints to the two generated functions (which contain
4063 `cyg_profile` in their names, hence the helper's name).
4065 To use the LTTng-UST function tracing helper, the source files to
4066 instrument must be built using the `-finstrument-functions` compiler
4069 There are two versions of the LTTng-UST function tracing helper:
4071 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4072 that you should only use when it can be _guaranteed_ that the
4073 complete event stream is recorded without any lost event record.
4074 Any kind of duplicate information is left out.
4076 Assuming no event record is lost, having only the function addresses on
4077 entry is enough to create a call graph, since an event record always
4078 contains the ID of the CPU that generated it.
4080 You can use a tool like man:addr2line(1) to convert function addresses
4081 back to source file names and line numbers.
4083 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4084 which also works in use cases where event records might get discarded or
4085 not recorded from application startup.
4086 In these cases, the trace analyzer needs more information to be
4087 able to reconstruct the program flow.
4089 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4090 points of this helper.
4092 All the tracepoints that this helper provides have the
4093 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4095 TIP: It's sometimes a good idea to limit the number of source files that
4096 you compile with the `-finstrument-functions` option to prevent LTTng
4097 from writing an excessive amount of trace data at run time. When using
4098 man:gcc(1), you can use the
4099 `-finstrument-functions-exclude-function-list` option to avoid
4100 instrument entries and exits of specific function names.
4105 ==== Instrument the dynamic linker
4107 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4108 man:dlopen(3) and man:dlclose(3) function calls.
4110 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4115 [[java-application]]
4116 === User space Java agent
4118 You can instrument any Java application which uses one of the following
4121 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4122 (JUL) core logging facilities.
4123 * http://logging.apache.org/log4j/1.2/[**Apache log4j{nbsp}1.2**], since
4124 LTTng{nbsp}2.6. Note that Apache Log4j{nbsp}2 is not supported.
4127 .LTTng-UST Java agent imported by a Java application.
4128 image::java-app.png[]
4130 Note that the methods described below are new in LTTng{nbsp}{revision}.
4131 Previous LTTng versions use another technique.
4133 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4134 and https://ci.lttng.org/[continuous integration], thus this version is
4135 directly supported. However, the LTTng-UST Java agent is also tested
4136 with OpenJDK{nbsp}7.
4141 ==== Use the LTTng-UST Java agent for `java.util.logging`
4143 To use the LTTng-UST Java agent in a Java application which uses
4144 `java.util.logging` (JUL):
4146 . In the Java application's source code, import the LTTng-UST
4147 log handler package for `java.util.logging`:
4152 import org.lttng.ust.agent.jul.LttngLogHandler;
4156 . Create an LTTng-UST JUL log handler:
4161 Handler lttngUstLogHandler = new LttngLogHandler();
4165 . Add this handler to the JUL loggers which should emit LTTng events:
4170 Logger myLogger = Logger.getLogger("some-logger");
4172 myLogger.addHandler(lttngUstLogHandler);
4176 . Use `java.util.logging` log statements and configuration as usual.
4177 The loggers with an attached LTTng-UST log handler can emit
4180 . Before exiting the application, remove the LTTng-UST log handler from
4181 the loggers attached to it and call its `close()` method:
4186 myLogger.removeHandler(lttngUstLogHandler);
4187 lttngUstLogHandler.close();
4191 This is not strictly necessary, but it is recommended for a clean
4192 disposal of the handler's resources.
4194 . Include the LTTng-UST Java agent's common and JUL-specific JAR files,
4195 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4197 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4198 path] when you build the Java application.
4200 The JAR files are typically located in dir:{/usr/share/java}.
4202 IMPORTANT: The LTTng-UST Java agent must be
4203 <<installing-lttng,installed>> for the logging framework your
4206 .Use the LTTng-UST Java agent for `java.util.logging`.
4211 import java.io.IOException;
4212 import java.util.logging.Handler;
4213 import java.util.logging.Logger;
4214 import org.lttng.ust.agent.jul.LttngLogHandler;
4218 private static final int answer = 42;
4220 public static void main(String[] argv) throws Exception
4223 Logger logger = Logger.getLogger("jello");
4225 // Create an LTTng-UST log handler
4226 Handler lttngUstLogHandler = new LttngLogHandler();
4228 // Add the LTTng-UST log handler to our logger
4229 logger.addHandler(lttngUstLogHandler);
4232 logger.info("some info");
4233 logger.warning("some warning");
4235 logger.finer("finer information; the answer is " + answer);
4237 logger.severe("error!");
4239 // Not mandatory, but cleaner
4240 logger.removeHandler(lttngUstLogHandler);
4241 lttngUstLogHandler.close();
4250 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4253 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4254 <<enabling-disabling-events,create an event rule>> matching the
4255 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4260 $ lttng enable-event --jul jello
4264 Run the compiled class:
4268 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4271 <<basic-tracing-session-control,Stop tracing>> and inspect the
4281 In the resulting trace, an <<event,event record>> generated by a Java
4282 application using `java.util.logging` is named `lttng_jul:event` and
4283 has the following fields:
4286 Log record's message.
4292 Name of the class in which the log statement was executed.
4295 Name of the method in which the log statement was executed.
4298 Logging time (timestamp in milliseconds).
4301 Log level integer value.
4304 ID of the thread in which the log statement was executed.
4306 You can use the opt:lttng-enable-event(1):--loglevel or
4307 opt:lttng-enable-event(1):--loglevel-only option of the
4308 man:lttng-enable-event(1) command to target a range of JUL log levels
4309 or a specific JUL log level.
4314 ==== Use the LTTng-UST Java agent for Apache log4j
4316 To use the LTTng-UST Java agent in a Java application which uses
4317 Apache log4j{nbsp}1.2:
4319 . In the Java application's source code, import the LTTng-UST
4320 log appender package for Apache log4j:
4325 import org.lttng.ust.agent.log4j.LttngLogAppender;
4329 . Create an LTTng-UST log4j log appender:
4334 Appender lttngUstLogAppender = new LttngLogAppender();
4338 . Add this appender to the log4j loggers which should emit LTTng events:
4343 Logger myLogger = Logger.getLogger("some-logger");
4345 myLogger.addAppender(lttngUstLogAppender);
4349 . Use Apache log4j log statements and configuration as usual. The
4350 loggers with an attached LTTng-UST log appender can emit LTTng events.
4352 . Before exiting the application, remove the LTTng-UST log appender from
4353 the loggers attached to it and call its `close()` method:
4358 myLogger.removeAppender(lttngUstLogAppender);
4359 lttngUstLogAppender.close();
4363 This is not strictly necessary, but it is recommended for a clean
4364 disposal of the appender's resources.
4366 . Include the LTTng-UST Java agent's common and log4j-specific JAR
4367 files, path:{lttng-ust-agent-common.jar} and
4368 path:{lttng-ust-agent-log4j.jar}, in the
4369 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4370 path] when you build the Java application.
4372 The JAR files are typically located in dir:{/usr/share/java}.
4374 IMPORTANT: The LTTng-UST Java agent must be
4375 <<installing-lttng,installed>> for the logging framework your
4378 .Use the LTTng-UST Java agent for Apache log4j.
4383 import org.apache.log4j.Appender;
4384 import org.apache.log4j.Logger;
4385 import org.lttng.ust.agent.log4j.LttngLogAppender;
4389 private static final int answer = 42;
4391 public static void main(String[] argv) throws Exception
4394 Logger logger = Logger.getLogger("jello");
4396 // Create an LTTng-UST log appender
4397 Appender lttngUstLogAppender = new LttngLogAppender();
4399 // Add the LTTng-UST log appender to our logger
4400 logger.addAppender(lttngUstLogAppender);
4403 logger.info("some info");
4404 logger.warn("some warning");
4406 logger.debug("debug information; the answer is " + answer);
4408 logger.fatal("error!");
4410 // Not mandatory, but cleaner
4411 logger.removeAppender(lttngUstLogAppender);
4412 lttngUstLogAppender.close();
4418 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4423 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4426 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4427 <<enabling-disabling-events,create an event rule>> matching the
4428 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4433 $ lttng enable-event --log4j jello
4437 Run the compiled class:
4441 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4444 <<basic-tracing-session-control,Stop tracing>> and inspect the
4454 In the resulting trace, an <<event,event record>> generated by a Java
4455 application using log4j is named `lttng_log4j:event` and
4456 has the following fields:
4459 Log record's message.
4465 Name of the class in which the log statement was executed.
4468 Name of the method in which the log statement was executed.
4471 Name of the file in which the executed log statement is located.
4474 Line number at which the log statement was executed.
4480 Log level integer value.
4483 Name of the Java thread in which the log statement was executed.
4485 You can use the opt:lttng-enable-event(1):--loglevel or
4486 opt:lttng-enable-event(1):--loglevel-only option of the
4487 man:lttng-enable-event(1) command to target a range of Apache log4j log levels
4488 or a specific log4j log level.
4492 [[java-application-context]]
4493 ==== Provide application-specific context fields in a Java application
4495 A Java application-specific context field is a piece of state provided
4496 by the application which <<adding-context,you can add>>, using the
4497 man:lttng-add-context(1) command, to each <<event,event record>>
4498 produced by the log statements of this application.
4500 For example, a given object might have a current request ID variable.
4501 You can create a context information retriever for this object and
4502 assign a name to this current request ID. You can then, using the
4503 man:lttng-add-context(1) command, add this context field by name to
4504 the JUL or log4j <<channel,channel>>.
4506 To provide application-specific context fields in a Java application:
4508 . In the Java application's source code, import the LTTng-UST
4509 Java agent context classes and interfaces:
4514 import org.lttng.ust.agent.context.ContextInfoManager;
4515 import org.lttng.ust.agent.context.IContextInfoRetriever;
4519 . Create a context information retriever class, that is, a class which
4520 implements the `IContextInfoRetriever` interface:
4525 class MyContextInfoRetriever implements IContextInfoRetriever
4528 public Object retrieveContextInfo(String key)
4530 if (key.equals("intCtx")) {
4532 } else if (key.equals("strContext")) {
4533 return "context value!";
4542 This `retrieveContextInfo()` method is the only member of the
4543 `IContextInfoRetriever` interface. Its role is to return the current
4544 value of a state by name to create a context field. The names of the
4545 context fields and which state variables they return depends on your
4548 All primitive types and objects are supported as context fields.
4549 When `retrieveContextInfo()` returns an object, the context field
4550 serializer calls its `toString()` method to add a string field to
4551 event records. The method can also return `null`, which means that
4552 no context field is available for the required name.
4554 . Register an instance of your context information retriever class to
4555 the context information manager singleton:
4560 IContextInfoRetriever cir = new MyContextInfoRetriever();
4561 ContextInfoManager cim = ContextInfoManager.getInstance();
4562 cim.registerContextInfoRetriever("retrieverName", cir);
4566 . Before exiting the application, remove your context information
4567 retriever from the context information manager singleton:
4572 ContextInfoManager cim = ContextInfoManager.getInstance();
4573 cim.unregisterContextInfoRetriever("retrieverName");
4577 This is not strictly necessary, but it is recommended for a clean
4578 disposal of some manager's resources.
4580 . Build your Java application with LTTng-UST Java agent support as
4581 usual, following the procedure for either the <<jul,JUL>> or
4582 <<log4j,Apache log4j>> framework.
4585 .Provide application-specific context fields in a Java application.
4590 import java.util.logging.Handler;
4591 import java.util.logging.Logger;
4592 import org.lttng.ust.agent.jul.LttngLogHandler;
4593 import org.lttng.ust.agent.context.ContextInfoManager;
4594 import org.lttng.ust.agent.context.IContextInfoRetriever;
4598 // Our context information retriever class
4599 private static class MyContextInfoRetriever
4600 implements IContextInfoRetriever
4603 public Object retrieveContextInfo(String key) {
4604 if (key.equals("intCtx")) {
4606 } else if (key.equals("strContext")) {
4607 return "context value!";
4614 private static final int answer = 42;
4616 public static void main(String args[]) throws Exception
4618 // Get the context information manager instance
4619 ContextInfoManager cim = ContextInfoManager.getInstance();
4621 // Create and register our context information retriever
4622 IContextInfoRetriever cir = new MyContextInfoRetriever();
4623 cim.registerContextInfoRetriever("myRetriever", cir);
4626 Logger logger = Logger.getLogger("jello");
4628 // Create an LTTng-UST log handler
4629 Handler lttngUstLogHandler = new LttngLogHandler();
4631 // Add the LTTng-UST log handler to our logger
4632 logger.addHandler(lttngUstLogHandler);
4635 logger.info("some info");
4636 logger.warning("some warning");
4638 logger.finer("finer information; the answer is " + answer);
4640 logger.severe("error!");
4642 // Not mandatory, but cleaner
4643 logger.removeHandler(lttngUstLogHandler);
4644 lttngUstLogHandler.close();
4645 cim.unregisterContextInfoRetriever("myRetriever");
4654 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4657 <<creating-destroying-tracing-sessions,Create a tracing session>>
4658 and <<enabling-disabling-events,create an event rule>> matching the
4664 $ lttng enable-event --jul jello
4667 <<adding-context,Add the application-specific context fields>> to the
4672 $ lttng add-context --jul --type='$app.myRetriever:intCtx'
4673 $ lttng add-context --jul --type='$app.myRetriever:strContext'
4676 <<basic-tracing-session-control,Start tracing>>:
4683 Run the compiled class:
4687 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4690 <<basic-tracing-session-control,Stop tracing>> and inspect the
4702 [[python-application]]
4703 === User space Python agent
4705 You can instrument a Python{nbsp}2 or Python{nbsp}3 application which
4707 https://docs.python.org/3/library/logging.html[`logging`] package.
4709 Each log statement emits an LTTng event once the
4710 application module imports the
4711 <<lttng-ust-agents,LTTng-UST Python agent>> package.
4714 .A Python application importing the LTTng-UST Python agent.
4715 image::python-app.png[]
4717 To use the LTTng-UST Python agent:
4719 . In the Python application's source code, import the LTTng-UST Python
4729 The LTTng-UST Python agent automatically adds its logging handler to the
4730 root logger at import time.
4732 Any log statement that the application executes before this import does
4733 not emit an LTTng event.
4735 IMPORTANT: The LTTng-UST Python agent must be
4736 <<installing-lttng,installed>>.
4738 . Use log statements and logging configuration as usual.
4739 Since the LTTng-UST Python agent adds a handler to the _root_
4740 logger, you can trace any log statement from any logger.
4742 .Use the LTTng-UST Python agent.
4753 logging.basicConfig()
4754 logger = logging.getLogger('my-logger')
4757 logger.debug('debug message')
4758 logger.info('info message')
4759 logger.warn('warn message')
4760 logger.error('error message')
4761 logger.critical('critical message')
4765 if __name__ == '__main__':
4769 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
4770 logging handler which prints to the standard error stream, is not
4771 strictly required for LTTng-UST tracing to work, but in versions of
4772 Python preceding{nbsp}3.2, you could see a warning message which indicates
4773 that no handler exists for the logger `my-logger`.
4775 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4776 <<enabling-disabling-events,create an event rule>> matching the
4777 `my-logger` Python logger, and <<basic-tracing-session-control,start
4783 $ lttng enable-event --python my-logger
4787 Run the Python script:
4794 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4804 In the resulting trace, an <<event,event record>> generated by a Python
4805 application is named `lttng_python:event` and has the following fields:
4808 Logging time (string).
4811 Log record's message.
4817 Name of the function in which the log statement was executed.
4820 Line number at which the log statement was executed.
4823 Log level integer value.
4826 ID of the Python thread in which the log statement was executed.
4829 Name of the Python thread in which the log statement was executed.
4831 You can use the opt:lttng-enable-event(1):--loglevel or
4832 opt:lttng-enable-event(1):--loglevel-only option of the
4833 man:lttng-enable-event(1) command to target a range of Python log levels
4834 or a specific Python log level.
4836 When an application imports the LTTng-UST Python agent, the agent tries
4837 to register to a <<lttng-sessiond,session daemon>>. Note that you must
4838 <<start-sessiond,start the session daemon>> _before_ you run the Python
4839 application. If a session daemon is found, the agent tries to register
4840 to it during five seconds, after which the application continues
4841 without LTTng tracing support. You can override this timeout value with
4842 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
4845 If the session daemon stops while a Python application with an imported
4846 LTTng-UST Python agent runs, the agent retries to connect and to
4847 register to a session daemon every three seconds. You can override this
4848 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
4853 [[proc-lttng-logger-abi]]
4856 The `lttng-tracer` Linux kernel module, part of
4857 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
4858 path:{/proc/lttng-logger} when it's loaded. Any application can write
4859 text data to this file to emit an LTTng event.
4862 .An application writes to the LTTng logger file to emit an LTTng event.
4863 image::lttng-logger.png[]
4865 The LTTng logger is the quickest method--not the most efficient,
4866 however--to add instrumentation to an application. It is designed
4867 mostly to instrument shell scripts:
4871 $ echo "Some message, some $variable" > /proc/lttng-logger
4874 Any event that the LTTng logger emits is named `lttng_logger` and
4875 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
4876 other instrumentation points in the kernel tracing domain, **any Unix
4877 user** can <<enabling-disabling-events,create an event rule>> which
4878 matches its event name, not only the root user or users in the
4879 <<tracing-group,tracing group>>.
4881 To use the LTTng logger:
4883 * From any application, write text data to the path:{/proc/lttng-logger}
4886 The `msg` field of `lttng_logger` event records contains the
4889 NOTE: The maximum message length of an LTTng logger event is
4890 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
4891 than one event to contain the remaining data.
4893 You should not use the LTTng logger to trace a user application which
4894 can be instrumented in a more efficient way, namely:
4896 * <<c-application,C and $$C++$$ applications>>.
4897 * <<java-application,Java applications>>.
4898 * <<python-application,Python applications>>.
4900 .Use the LTTng logger.
4905 echo 'Hello, World!' > /proc/lttng-logger
4907 df --human-readable --print-type / > /proc/lttng-logger
4910 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4911 <<enabling-disabling-events,create an event rule>> matching the
4912 `lttng_logger` Linux kernel tracepoint, and
4913 <<basic-tracing-session-control,start tracing>>:
4918 $ lttng enable-event --kernel lttng_logger
4922 Run the Bash script:
4929 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
4940 [[instrumenting-linux-kernel]]
4941 === LTTng kernel tracepoints
4943 NOTE: This section shows how to _add_ instrumentation points to the
4944 Linux kernel. The kernel's subsystems are already thoroughly
4945 instrumented at strategic places for LTTng when you
4946 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
4950 There are two methods to instrument the Linux kernel:
4952 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
4953 tracepoint which uses the `TRACE_EVENT()` API.
4955 Choose this if you want to instrumentation a Linux kernel tree with an
4956 instrumentation point compatible with ftrace, perf, and SystemTap.
4958 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
4959 instrument an out-of-tree kernel module.
4961 Choose this if you don't need ftrace, perf, or SystemTap support.
4965 [[linux-add-lttng-layer]]
4966 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
4968 This section shows how to add an LTTng layer to existing ftrace
4969 instrumentation using the `TRACE_EVENT()` API.
4971 This section does not document the `TRACE_EVENT()` macro. You can
4972 read the following articles to learn more about this API:
4974 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part{nbsp}1)]
4975 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part{nbsp}2)]
4976 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part{nbsp}3)]
4978 The following procedure assumes that your ftrace tracepoints are
4979 correctly defined in their own header and that they are created in
4980 one source file using the `CREATE_TRACE_POINTS` definition.
4982 To add an LTTng layer over an existing ftrace tracepoint:
4984 . Make sure the following kernel configuration options are
4990 * `CONFIG_HIGH_RES_TIMERS`
4991 * `CONFIG_TRACEPOINTS`
4994 . Build the Linux source tree with your custom ftrace tracepoints.
4995 . Boot the resulting Linux image on your target system.
4997 Confirm that the tracepoints exist by looking for their names in the
4998 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
4999 is your subsystem's name.
5001 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5006 $ cd $(mktemp -d) &&
5007 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.11.tar.bz2 &&
5008 tar -xf lttng-modules-latest-2.11.tar.bz2 &&
5009 cd lttng-modules-2.11.*
5013 . In dir:{instrumentation/events/lttng-module}, relative to the root
5014 of the LTTng-modules source tree, create a header file named
5015 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5016 LTTng-modules tracepoint definitions using the LTTng-modules
5019 Start with this template:
5023 .path:{instrumentation/events/lttng-module/my_subsys.h}
5026 #define TRACE_SYSTEM my_subsys
5028 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5029 #define _LTTNG_MY_SUBSYS_H
5031 #include "../../../probes/lttng-tracepoint-event.h"
5032 #include <linux/tracepoint.h>
5034 LTTNG_TRACEPOINT_EVENT(
5036 * Format is identical to TRACE_EVENT()'s version for the three
5037 * following macro parameters:
5040 TP_PROTO(int my_int, const char *my_string),
5041 TP_ARGS(my_int, my_string),
5043 /* LTTng-modules specific macros */
5045 ctf_integer(int, my_int_field, my_int)
5046 ctf_string(my_bar_field, my_bar)
5050 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5052 #include "../../../probes/define_trace.h"
5056 The entries in the `TP_FIELDS()` section are the list of fields for the
5057 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5058 ftrace's `TRACE_EVENT()` macro.
5060 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5061 complete description of the available `ctf_*()` macros.
5063 . Create the LTTng-modules probe's kernel module C source file,
5064 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5069 .path:{probes/lttng-probe-my-subsys.c}
5071 #include <linux/module.h>
5072 #include "../lttng-tracer.h"
5075 * Build-time verification of mismatch between mainline
5076 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5077 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5079 #include <trace/events/my_subsys.h>
5081 /* Create LTTng tracepoint probes */
5082 #define LTTNG_PACKAGE_BUILD
5083 #define CREATE_TRACE_POINTS
5084 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5086 #include "../instrumentation/events/lttng-module/my_subsys.h"
5088 MODULE_LICENSE("GPL and additional rights");
5089 MODULE_AUTHOR("Your name <your-email>");
5090 MODULE_DESCRIPTION("LTTng my_subsys probes");
5091 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5092 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5093 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5094 LTTNG_MODULES_EXTRAVERSION);
5098 . Edit path:{probes/KBuild} and add your new kernel module object
5099 next to the existing ones:
5103 .path:{probes/KBuild}
5107 obj-m += lttng-probe-module.o
5108 obj-m += lttng-probe-power.o
5110 obj-m += lttng-probe-my-subsys.o
5116 . Build and install the LTTng kernel modules:
5121 $ make KERNELDIR=/path/to/linux
5122 # make modules_install && depmod -a
5126 Replace `/path/to/linux` with the path to the Linux source tree where
5127 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
5129 Note that you can also use the
5130 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5131 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5132 C code that need to be executed before the event fields are recorded.
5134 The best way to learn how to use the previous LTTng-modules macros is to
5135 inspect the existing LTTng-modules tracepoint definitions in the
5136 dir:{instrumentation/events/lttng-module} header files. Compare them
5137 with the Linux kernel mainline versions in the
5138 dir:{include/trace/events} directory of the Linux source tree.
5142 [[lttng-tracepoint-event-code]]
5143 ===== Use custom C code to access the data for tracepoint fields
5145 Although we recommended to always use the
5146 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5147 the arguments and fields of an LTTng-modules tracepoint when possible,
5148 sometimes you need a more complex process to access the data that the
5149 tracer records as event record fields. In other words, you need local
5150 variables and multiple C{nbsp}statements instead of simple
5151 argument-based expressions that you pass to the
5152 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5154 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5155 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5156 a block of C{nbsp}code to be executed before LTTng records the fields.
5157 The structure of this macro is:
5160 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5162 LTTNG_TRACEPOINT_EVENT_CODE(
5164 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5165 * version for the following three macro parameters:
5168 TP_PROTO(int my_int, const char *my_string),
5169 TP_ARGS(my_int, my_string),
5171 /* Declarations of custom local variables */
5174 unsigned long b = 0;
5175 const char *name = "(undefined)";
5176 struct my_struct *my_struct;
5180 * Custom code which uses both tracepoint arguments
5181 * (in TP_ARGS()) and local variables (in TP_locvar()).
5183 * Local variables are actually members of a structure pointed
5184 * to by the special variable tp_locvar.
5188 tp_locvar->a = my_int + 17;
5189 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5190 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5191 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5192 put_my_struct(tp_locvar->my_struct);
5201 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5202 * version for this, except that tp_locvar members can be
5203 * used in the argument expression parameters of
5204 * the ctf_*() macros.
5207 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5208 ctf_integer(int, my_struct_a, tp_locvar->a)
5209 ctf_string(my_string_field, my_string)
5210 ctf_string(my_struct_name, tp_locvar->name)
5215 IMPORTANT: The C code defined in `TP_code()` must not have any side
5216 effects when executed. In particular, the code must not allocate
5217 memory or get resources without deallocating this memory or putting
5218 those resources afterwards.
5221 [[instrumenting-linux-kernel-tracing]]
5222 ==== Load and unload a custom probe kernel module
5224 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5225 kernel module>> in the kernel before it can emit LTTng events.
5227 To load the default probe kernel modules and a custom probe kernel
5230 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5231 probe modules to load when starting a root <<lttng-sessiond,session
5235 .Load the `my_subsys`, `usb`, and the default probe modules.
5239 # lttng-sessiond --extra-kmod-probes=my_subsys,usb
5244 You only need to pass the subsystem name, not the whole kernel module
5247 To load _only_ a given custom probe kernel module:
5249 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5250 modules to load when starting a root session daemon:
5253 .Load only the `my_subsys` and `usb` probe modules.
5257 # lttng-sessiond --kmod-probes=my_subsys,usb
5262 To confirm that a probe module is loaded:
5269 $ lsmod | grep lttng_probe_usb
5273 To unload the loaded probe modules:
5275 * Kill the session daemon with `SIGTERM`:
5280 # pkill lttng-sessiond
5284 You can also use man:modprobe(8)'s `--remove` option if the session
5285 daemon terminates abnormally.
5288 [[controlling-tracing]]
5291 Once an application or a Linux kernel is
5292 <<instrumenting,instrumented>> for LTTng tracing,
5295 This section is divided in topics on how to use the various
5296 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5297 command-line tool>>, to _control_ the LTTng daemons and tracers.
5299 NOTE: In the following subsections, we refer to an man:lttng(1) command
5300 using its man page name. For example, instead of _Run the `create`
5301 command to..._, we use _Run the man:lttng-create(1) command to..._.
5305 === Start a session daemon
5307 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5308 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5311 You will see the following error when you run a command while no session
5315 Error: No session daemon is available
5318 The only command that automatically runs a session daemon is
5319 man:lttng-create(1), which you use to
5320 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5321 this is most of the time the first operation that you do, sometimes it's
5322 not. Some examples are:
5324 * <<list-instrumentation-points,List the available instrumentation points>>.
5325 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5327 [[tracing-group]] Each Unix user must have its own running session
5328 daemon to trace user applications. The session daemon that the root user
5329 starts is the only one allowed to control the LTTng kernel tracer. Users
5330 that are part of the _tracing group_ can control the root session
5331 daemon. The default tracing group name is `tracing`; you can set it to
5332 something else with the opt:lttng-sessiond(8):--group option when you
5333 start the root session daemon.
5335 To start a user session daemon:
5337 * Run man:lttng-sessiond(8):
5342 $ lttng-sessiond --daemonize
5346 To start the root session daemon:
5348 * Run man:lttng-sessiond(8) as the root user:
5353 # lttng-sessiond --daemonize
5357 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5358 start the session daemon in foreground.
5360 To stop a session daemon, use man:kill(1) on its process ID (standard
5363 Note that some Linux distributions could manage the LTTng session daemon
5364 as a service. In this case, you should use the service manager to
5365 start, restart, and stop session daemons.
5368 [[creating-destroying-tracing-sessions]]
5369 === Create and destroy a tracing session
5371 Almost all the LTTng control operations happen in the scope of
5372 a <<tracing-session,tracing session>>, which is the dialogue between the
5373 <<lttng-sessiond,session daemon>> and you.
5375 To create a tracing session with a generated name:
5377 * Use the man:lttng-create(1) command:
5386 The created tracing session's name is `auto` followed by the
5389 To create a tracing session with a specific name:
5391 * Use the optional argument of the man:lttng-create(1) command:
5396 $ lttng create my-session
5400 Replace `my-session` with the specific tracing session name.
5402 LTTng appends the creation date to the created tracing session's name.
5404 LTTng writes the traces of a tracing session in
5405 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5406 name of the tracing session. Note that the env:LTTNG_HOME environment
5407 variable defaults to `$HOME` if not set.
5409 To output LTTng traces to a non-default location:
5411 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5416 $ lttng create my-session --output=/tmp/some-directory
5420 You may create as many tracing sessions as you wish.
5422 To list all the existing tracing sessions for your Unix user:
5424 * Use the man:lttng-list(1) command:
5433 When you create a tracing session, it is set as the _current tracing
5434 session_. The following man:lttng(1) commands operate on the current
5435 tracing session when you don't specify one:
5437 [role="list-3-cols"]
5438 * man:lttng-add-context(1)
5439 * man:lttng-destroy(1)
5440 * man:lttng-disable-channel(1)
5441 * man:lttng-disable-event(1)
5442 * man:lttng-disable-rotation(1)
5443 * man:lttng-enable-channel(1)
5444 * man:lttng-enable-event(1)
5445 * man:lttng-enable-rotation(1)
5447 * man:lttng-regenerate(1)
5448 * man:lttng-rotate(1)
5450 * man:lttng-snapshot(1)
5451 * man:lttng-start(1)
5452 * man:lttng-status(1)
5454 * man:lttng-track(1)
5455 * man:lttng-untrack(1)
5458 To change the current tracing session:
5460 * Use the man:lttng-set-session(1) command:
5465 $ lttng set-session new-session
5469 Replace `new-session` by the name of the new current tracing session.
5471 When you are done tracing in a given tracing session, you can destroy
5472 it. This operation frees the resources taken by the tracing session
5473 to destroy; it does not destroy the trace data that LTTng wrote for
5474 this tracing session.
5476 To destroy the current tracing session:
5478 * Use the man:lttng-destroy(1) command:
5487 The man:lttng-destroy(1) command also runs the man:lttng-stop(1)
5488 command implicitly (see <<basic-tracing-session-control,Start and stop a
5489 tracing session>>). You need to stop tracing to make LTTng flush the
5490 remaining trace data and make the trace readable.
5493 [[list-instrumentation-points]]
5494 === List the available instrumentation points
5496 The <<lttng-sessiond,session daemon>> can query the running instrumented
5497 user applications and the Linux kernel to get a list of available
5498 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5499 they are tracepoints and system calls. For the user space tracing
5500 domain, they are tracepoints. For the other tracing domains, they are
5503 To list the available instrumentation points:
5505 * Use the man:lttng-list(1) command with the requested tracing domain's
5509 * opt:lttng-list(1):--kernel: Linux kernel tracepoints (your Unix user
5510 must be a root user, or it must be a member of the
5511 <<tracing-group,tracing group>>).
5512 * opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall: Linux
5513 kernel system calls (your Unix user must be a root user, or it must be
5514 a member of the tracing group).
5515 * opt:lttng-list(1):--userspace: user space tracepoints.
5516 * opt:lttng-list(1):--jul: `java.util.logging` loggers.
5517 * opt:lttng-list(1):--log4j: Apache log4j loggers.
5518 * opt:lttng-list(1):--python: Python loggers.
5521 .List the available user space tracepoints.
5525 $ lttng list --userspace
5529 .List the available Linux kernel system call tracepoints.
5533 $ lttng list --kernel --syscall
5538 [[enabling-disabling-events]]
5539 === Create and enable an event rule
5541 Once you <<creating-destroying-tracing-sessions,create a tracing
5542 session>>, you can create <<event,event rules>> with the
5543 man:lttng-enable-event(1) command.
5545 You specify each condition with a command-line option. The available
5546 condition arguments are shown in the following table.
5548 [role="growable",cols="asciidoc,asciidoc,default"]
5549 .Condition command-line arguments for the man:lttng-enable-event(1) command.
5551 |Argument |Description |Applicable tracing domains
5557 . +--probe=__ADDR__+
5558 . +--function=__ADDR__+
5559 . +--userspace-probe=__PATH__:__SYMBOL__+
5560 . +--userspace-probe=sdt:__PATH__:__PROVIDER__:__NAME__+
5563 Instead of using the default _tracepoint_ instrumentation type, use:
5565 . A Linux system call (entry and exit).
5566 . A Linux https://lwn.net/Articles/132196/[kprobe] (symbol or address).
5567 . The entry and return points of a Linux function (symbol or address).
5568 . The entry point of a user application or library function (path to
5569 application/library and symbol).
5570 . A https://www.sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps[SystemTap
5571 Statically Defined Tracing] (USDT) probe (path to application/library,
5572 provider and probe names).
5576 |First positional argument.
5579 Tracepoint or system call name.
5581 With the opt:lttng-enable-event(1):--probe,
5582 opt:lttng-enable-event(1):--function, and
5583 opt:lttng-enable-event(1):--userspace-probe options, this is a custom
5584 name given to the event rule. With the JUL, log4j, and Python domains,
5585 this is a logger name.
5587 With a tracepoint, logger, or system call name, you can use the special
5588 `*` globbing character to match anything (for example, `sched_*`,
5596 . +--loglevel=__LEVEL__+
5597 . +--loglevel-only=__LEVEL__+
5600 . Match only tracepoints or log statements with a logging level at
5601 least as severe as +__LEVEL__+.
5602 . Match only tracepoints or log statements with a logging level
5603 equal to +__LEVEL__+.
5605 See man:lttng-enable-event(1) for the list of available logging level
5608 |User space, JUL, log4j, and Python.
5610 |+--exclude=__EXCLUSIONS__+
5613 When you use a `*` character at the end of the tracepoint or logger
5614 name (first positional argument), exclude the specific names in the
5615 comma-delimited list +__EXCLUSIONS__+.
5618 User space, JUL, log4j, and Python.
5620 |+--filter=__EXPR__+
5623 Match only events which satisfy the expression +__EXPR__+.
5625 See man:lttng-enable-event(1) to learn more about the syntax of a
5632 You attach an event rule to a <<channel,channel>> on creation. If you do
5633 not specify the channel with the opt:lttng-enable-event(1):--channel
5634 option, and if the event rule to create is the first in its
5635 <<domain,tracing domain>> for a given tracing session, then LTTng
5636 creates a _default channel_ for you. This default channel is reused in
5637 subsequent invocations of the man:lttng-enable-event(1) command for the
5638 same tracing domain.
5640 An event rule is always enabled at creation time.
5642 The following examples show how you can combine the previous
5643 command-line options to create simple to more complex event rules.
5645 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5649 $ lttng enable-event --kernel sched_switch
5653 .Create an event rule matching four Linux kernel system calls (default channel).
5657 $ lttng enable-event --kernel --syscall open,write,read,close
5661 .Create event rules matching tracepoints with filter expressions (default channel).
5665 $ lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5670 $ lttng enable-event --kernel --all \
5671 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5676 $ lttng enable-event --jul my_logger \
5677 --filter='$app.retriever:cur_msg_id > 3'
5680 IMPORTANT: Make sure to always quote the filter string when you
5681 use man:lttng(1) from a shell.
5684 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5688 $ lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5691 IMPORTANT: Make sure to always quote the wildcard character when you
5692 use man:lttng(1) from a shell.
5695 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5699 $ lttng enable-event --python my-app.'*' \
5700 --exclude='my-app.module,my-app.hello'
5704 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5708 $ lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5712 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5716 $ lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5720 .Create an event rule matching the `malloc` function entry in path:{/usr/lib/libc.so.6}:
5724 $ lttng enable-event --kernel --userspace-probe=/usr/lib/libc.so.6:malloc \
5729 .Create an event rule matching the `server`/`accept_request` https://www.sourceware.org/systemtap/wiki/AddingUserSpaceProbingToApps[USDT probe] in path:{/usr/bin/serv}:
5733 $ lttng enable-event --kernel --userspace-probe=sdt:serv:server:accept_request \
5734 server_accept_request
5738 The event rules of a given channel form a whitelist: as soon as an
5739 emitted event passes one of them, LTTng can record the event. For
5740 example, an event named `my_app:my_tracepoint` emitted from a user space
5741 tracepoint with a `TRACE_ERROR` log level passes both of the following
5746 $ lttng enable-event --userspace my_app:my_tracepoint
5747 $ lttng enable-event --userspace my_app:my_tracepoint \
5748 --loglevel=TRACE_INFO
5751 The second event rule is redundant: the first one includes
5755 [[disable-event-rule]]
5756 === Disable an event rule
5758 To disable an event rule that you <<enabling-disabling-events,created>>
5759 previously, use the man:lttng-disable-event(1) command. This command
5760 disables _all_ the event rules (of a given tracing domain and channel)
5761 which match an instrumentation point. The other conditions are not
5762 supported as of LTTng{nbsp}{revision}.
5764 The LTTng tracer does not record an emitted event which passes
5765 a _disabled_ event rule.
5767 .Disable an event rule matching a Python logger (default channel).
5771 $ lttng disable-event --python my-logger
5775 .Disable an event rule matching all `java.util.logging` loggers (default channel).
5779 $ lttng disable-event --jul '*'
5783 .Disable _all_ the event rules of the default channel.
5785 The opt:lttng-disable-event(1):--all-events option is not, like the
5786 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
5787 equivalent of the event name `*` (wildcard): it disables _all_ the event
5788 rules of a given channel.
5792 $ lttng disable-event --jul --all-events
5796 NOTE: You cannot delete an event rule once you create it.
5800 === Get the status of a tracing session
5802 To get the status of the current tracing session, that is, its
5803 parameters, its channels, event rules, and their attributes:
5805 * Use the man:lttng-status(1) command:
5815 To get the status of any tracing session:
5817 * Use the man:lttng-list(1) command with the tracing session's name:
5822 $ lttng list my-session
5826 Replace `my-session` with the desired tracing session's name.
5829 [[basic-tracing-session-control]]
5830 === Start and stop a tracing session
5832 Once you <<creating-destroying-tracing-sessions,create a tracing
5834 <<enabling-disabling-events,create one or more event rules>>,
5835 you can start and stop the tracers for this tracing session.
5837 To start tracing in the current tracing session:
5839 * Use the man:lttng-start(1) command:
5848 LTTng is very flexible: you can launch user applications before
5849 or after the you start the tracers. The tracers only record the events
5850 if they pass enabled event rules and if they occur while the tracers are
5853 To stop tracing in the current tracing session:
5855 * Use the man:lttng-stop(1) command:
5864 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
5865 records>> or lost sub-buffers since the last time you ran
5866 man:lttng-start(1), warnings are printed when you run the
5867 man:lttng-stop(1) command.
5869 IMPORTANT: You need to stop tracing to make LTTng flush the remaining
5870 trace data and make the trace readable. Note that the
5871 man:lttng-destroy(1) command (see
5872 <<creating-destroying-tracing-sessions,Create and destroy a tracing
5873 session>>) also runs the man:lttng-stop(1) command implicitly.
5876 [[enabling-disabling-channels]]
5877 === Create a channel
5879 Once you create a tracing session, you can create a <<channel,channel>>
5880 with the man:lttng-enable-channel(1) command.
5882 Note that LTTng automatically creates a default channel when, for a
5883 given <<domain,tracing domain>>, no channels exist and you
5884 <<enabling-disabling-events,create>> the first event rule. This default
5885 channel is named `channel0` and its attributes are set to reasonable
5886 values. Therefore, you only need to create a channel when you need
5887 non-default attributes.
5889 You specify each non-default channel attribute with a command-line
5890 option when you use the man:lttng-enable-channel(1) command. The
5891 available command-line options are:
5893 [role="growable",cols="asciidoc,asciidoc"]
5894 .Command-line options for the man:lttng-enable-channel(1) command.
5896 |Option |Description
5902 <<channel-overwrite-mode-vs-discard-mode,event record loss mode>> instead
5903 of the default _discard_ mode.
5905 |`--buffers-pid` (user space tracing domain only)
5908 Use the per-process <<channel-buffering-schemes,buffering scheme>>
5909 instead of the default per-user buffering scheme.
5911 |+--subbuf-size=__SIZE__+
5914 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
5915 either for each Unix user (default), or for each instrumented process.
5917 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5919 |+--num-subbuf=__COUNT__+
5922 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
5923 for each Unix user (default), or for each instrumented process.
5925 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
5927 |+--tracefile-size=__SIZE__+
5930 Set the maximum size of each trace file that this channel writes within
5931 a stream to +__SIZE__+ bytes instead of no maximum.
5933 See <<tracefile-rotation,Trace file count and size>>.
5935 |+--tracefile-count=__COUNT__+
5938 Limit the number of trace files that this channel creates to
5939 +__COUNT__+ channels instead of no limit.
5941 See <<tracefile-rotation,Trace file count and size>>.
5943 |+--switch-timer=__PERIODUS__+
5946 Set the <<channel-switch-timer,switch timer period>>
5947 to +__PERIODUS__+{nbsp}µs.
5949 |+--read-timer=__PERIODUS__+
5952 Set the <<channel-read-timer,read timer period>>
5953 to +__PERIODUS__+{nbsp}µs.
5955 |[[opt-blocking-timeout]]+--blocking-timeout=__TIMEOUTUS__+
5958 Set the timeout of user space applications which load LTTng-UST
5959 in blocking mode to +__TIMEOUTUS__+:
5962 Never block (non-blocking mode).
5965 Block forever until space is available in a sub-buffer to record
5968 __n__, a positive value::
5969 Wait for at most __n__ µs when trying to write into a sub-buffer.
5971 Note that, for this option to have any effect on an instrumented
5972 user space application, you need to run the application with a set
5973 env:LTTNG_UST_ALLOW_BLOCKING environment variable.
5975 |+--output=__TYPE__+ (Linux kernel tracing domain only)
5978 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
5982 You can only create a channel in the Linux kernel and user space
5983 <<domain,tracing domains>>: other tracing domains have their own channel
5984 created on the fly when <<enabling-disabling-events,creating event
5989 Because of a current LTTng limitation, you must create all channels
5990 _before_ you <<basic-tracing-session-control,start tracing>> in a given
5991 tracing session, that is, before the first time you run
5994 Since LTTng automatically creates a default channel when you use the
5995 man:lttng-enable-event(1) command with a specific tracing domain, you
5996 cannot, for example, create a Linux kernel event rule, start tracing,
5997 and then create a user space event rule, because no user space channel
5998 exists yet and it's too late to create one.
6000 For this reason, make sure to configure your channels properly
6001 before starting the tracers for the first time!
6004 The following examples show how you can combine the previous
6005 command-line options to create simple to more complex channels.
6007 .Create a Linux kernel channel with default attributes.
6011 $ lttng enable-channel --kernel my-channel
6015 .Create a user space channel with four sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6019 $ lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6020 --buffers-pid my-channel
6024 .[[blocking-timeout-example]]Create a default user space channel with an infinite blocking timeout.
6026 <<creating-destroying-tracing-sessions,Create a tracing-session>>,
6027 create the channel, <<enabling-disabling-events,create an event rule>>,
6028 and <<basic-tracing-session-control,start tracing>>:
6033 $ lttng enable-channel --userspace --blocking-timeout=inf blocking-channel
6034 $ lttng enable-event --userspace --channel=blocking-channel --all
6038 Run an application instrumented with LTTng-UST and allow it to block:
6042 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
6046 .Create a Linux kernel channel which rotates eight trace files of 4{nbsp}MiB each for each stream
6050 $ lttng enable-channel --kernel --tracefile-count=8 \
6051 --tracefile-size=4194304 my-channel
6055 .Create a user space channel in overwrite (or _flight recorder_) mode.
6059 $ lttng enable-channel --userspace --overwrite my-channel
6063 You can <<enabling-disabling-events,create>> the same event rule in
6064 two different channels:
6068 $ lttng enable-event --userspace --channel=my-channel app:tp
6069 $ lttng enable-event --userspace --channel=other-channel app:tp
6072 If both channels are enabled, when a tracepoint named `app:tp` is
6073 reached, LTTng records two events, one for each channel.
6077 === Disable a channel
6079 To disable a specific channel that you <<enabling-disabling-channels,created>>
6080 previously, use the man:lttng-disable-channel(1) command.
6082 .Disable a specific Linux kernel channel.
6086 $ lttng disable-channel --kernel my-channel
6090 The state of a channel precedes the individual states of event rules
6091 attached to it: event rules which belong to a disabled channel, even if
6092 they are enabled, are also considered disabled.
6096 === Add context fields to a channel
6098 Event record fields in trace files provide important information about
6099 events that occured previously, but sometimes some external context may
6100 help you solve a problem faster. Examples of context fields are:
6102 * The **process ID**, **thread ID**, **process name**, and
6103 **process priority** of the thread in which the event occurs.
6104 * The **hostname** of the system on which the event occurs.
6105 * The Linux kernel and user call stacks (since
6106 LTTng{nbsp}{revision}).
6107 * The current values of many possible **performance counters** using
6109 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6111 ** Branch instructions, misses, and loads.
6113 * Any context defined at the application level (supported for the
6114 JUL and log4j <<domain,tracing domains>>).
6116 To get the full list of available context fields, see
6117 `lttng add-context --list`. Some context fields are reserved for a
6118 specific <<domain,tracing domain>> (Linux kernel or user space).
6120 You add context fields to <<channel,channels>>. All the events
6121 that a channel with added context fields records contain those fields.
6123 To add context fields to one or all the channels of a given tracing
6126 * Use the man:lttng-add-context(1) command.
6128 .Add context fields to all the channels of the current tracing session.
6130 The following command line adds the virtual process identifier and
6131 the per-thread CPU cycles count fields to all the user space channels
6132 of the current tracing session.
6136 $ lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6140 .Add performance counter context fields by raw ID
6142 See man:lttng-add-context(1) for the exact format of the context field
6143 type, which is partly compatible with the format used in
6148 $ lttng add-context --userspace --type=perf:thread:raw:r0110:test
6149 $ lttng add-context --kernel --type=perf:cpu:raw:r0013c:x86unhalted
6153 .Add context fields to a specific channel.
6155 The following command line adds the thread identifier and user call
6156 stack context fields to the Linux kernel channel named `my-channel` in
6157 the current tracing session.
6161 $ lttng add-context --kernel --channel=my-channel \
6162 --type=tid --type=callstack-user
6166 .Add an application-specific context field to a specific channel.
6168 The following command line adds the `cur_msg_id` context field of the
6169 `retriever` context retriever for all the instrumented
6170 <<java-application,Java applications>> recording <<event,event records>>
6171 in the channel named `my-channel`:
6175 $ lttng add-context --kernel --channel=my-channel \
6176 --type='$app:retriever:cur_msg_id'
6179 IMPORTANT: Make sure to always quote the `$` character when you
6180 use man:lttng-add-context(1) from a shell.
6183 NOTE: You cannot remove context fields from a channel once you add it.
6188 === Track process IDs
6190 It's often useful to allow only specific process IDs (PIDs) to emit
6191 events. For example, you may wish to record all the system calls made by
6192 a given process (Ă la http://linux.die.net/man/1/strace[strace]).
6194 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6195 purpose. Both commands operate on a whitelist of process IDs. You _add_
6196 entries to this whitelist with the man:lttng-track(1) command and remove
6197 entries with the man:lttng-untrack(1) command. Any process which has one
6198 of the PIDs in the whitelist is allowed to emit LTTng events which pass
6199 an enabled <<event,event rule>>.
6201 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
6202 process with a given tracked ID exit and another process be given this
6203 ID, then the latter would also be allowed to emit events.
6205 .Track and untrack process IDs.
6207 For the sake of the following example, assume the target system has
6208 16{nbsp}possible PIDs.
6211 <<creating-destroying-tracing-sessions,create a tracing session>>,
6212 the whitelist contains all the possible PIDs:
6215 .All PIDs are tracked.
6216 image::track-all.png[]
6218 When the whitelist is full and you use the man:lttng-track(1) command to
6219 specify some PIDs to track, LTTng first clears the whitelist, then it
6220 tracks the specific PIDs. After:
6224 $ lttng track --pid=3,4,7,10,13
6230 .PIDs 3, 4, 7, 10, and 13 are tracked.
6231 image::track-3-4-7-10-13.png[]
6233 You can add more PIDs to the whitelist afterwards:
6237 $ lttng track --pid=1,15,16
6243 .PIDs 1, 15, and 16 are added to the whitelist.
6244 image::track-1-3-4-7-10-13-15-16.png[]
6246 The man:lttng-untrack(1) command removes entries from the PID tracker's
6247 whitelist. Given the previous example, the following command:
6251 $ lttng untrack --pid=3,7,10,13
6254 leads to this whitelist:
6257 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
6258 image::track-1-4-15-16.png[]
6260 LTTng can track all possible PIDs again using the
6261 opt:lttng-track(1):--all option:
6265 $ lttng track --pid --all
6268 The result is, again:
6271 .All PIDs are tracked.
6272 image::track-all.png[]
6275 .Track only specific PIDs
6277 A very typical use case with PID tracking is to start with an empty
6278 whitelist, then <<basic-tracing-session-control,start the tracers>>, and
6279 then add PIDs manually while tracers are active. You can accomplish this
6280 by using the opt:lttng-untrack(1):--all option of the
6281 man:lttng-untrack(1) command to clear the whitelist after you
6282 <<creating-destroying-tracing-sessions,create a tracing session>>:
6286 $ lttng untrack --pid --all
6292 .No PIDs are tracked.
6293 image::untrack-all.png[]
6295 If you trace with this whitelist configuration, the tracer records no
6296 events for this <<domain,tracing domain>> because no processes are
6297 tracked. You can use the man:lttng-track(1) command as usual to track
6298 specific PIDs, for example:
6302 $ lttng track --pid=6,11
6308 .PIDs 6 and 11 are tracked.
6309 image::track-6-11.png[]
6314 [[saving-loading-tracing-session]]
6315 === Save and load tracing session configurations
6317 Configuring a <<tracing-session,tracing session>> can be long. Some of
6318 the tasks involved are:
6320 * <<enabling-disabling-channels,Create channels>> with
6321 specific attributes.
6322 * <<adding-context,Add context fields>> to specific channels.
6323 * <<enabling-disabling-events,Create event rules>> with specific log
6324 level and filter conditions.
6326 If you use LTTng to solve real world problems, chances are you have to
6327 record events using the same tracing session setup over and over,
6328 modifying a few variables each time in your instrumented program
6329 or environment. To avoid constant tracing session reconfiguration,
6330 the man:lttng(1) command-line tool can save and load tracing session
6331 configurations to/from XML files.
6333 To save a given tracing session configuration:
6335 * Use the man:lttng-save(1) command:
6340 $ lttng save my-session
6344 Replace `my-session` with the name of the tracing session to save.
6346 LTTng saves tracing session configurations to
6347 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6348 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6349 the opt:lttng-save(1):--output-path option to change this destination
6352 LTTng saves all configuration parameters, for example:
6354 * The tracing session name.
6355 * The trace data output path.
6356 * The channels with their state and all their attributes.
6357 * The context fields you added to channels.
6358 * The event rules with their state, log level and filter conditions.
6360 To load a tracing session:
6362 * Use the man:lttng-load(1) command:
6367 $ lttng load my-session
6371 Replace `my-session` with the name of the tracing session to load.
6373 When LTTng loads a configuration, it restores your saved tracing session
6374 as if you just configured it manually.
6376 See man:lttng-load(1) for the complete list of command-line options. You
6377 can also save and load many sessions at a time, and decide in which
6378 directory to output the XML files.
6381 [[sending-trace-data-over-the-network]]
6382 === Send trace data over the network
6384 LTTng can send the recorded trace data to a remote system over the
6385 network instead of writing it to the local file system.
6387 To send the trace data over the network:
6389 . On the _remote_ system (which can also be the target system),
6390 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6399 . On the _target_ system, create a tracing session configured to
6400 send trace data over the network:
6405 $ lttng create my-session --set-url=net://remote-system
6409 Replace `remote-system` by the host name or IP address of the
6410 remote system. See man:lttng-create(1) for the exact URL format.
6412 . On the target system, use the man:lttng(1) command-line tool as usual.
6413 When tracing is active, the target's consumer daemon sends sub-buffers
6414 to the relay daemon running on the remote system instead of flushing
6415 them to the local file system. The relay daemon writes the received
6416 packets to the local file system.
6418 The relay daemon writes trace files to
6419 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6420 +__hostname__+ is the host name of the target system and +__session__+
6421 is the tracing session name. Note that the env:LTTNG_HOME environment
6422 variable defaults to `$HOME` if not set. Use the
6423 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6424 trace files to another base directory.
6429 === View events as LTTng emits them (noch:{LTTng} live)
6431 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6432 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6433 display events as LTTng emits them on the target system while tracing is
6436 The relay daemon creates a _tee_: it forwards the trace data to both
6437 the local file system and to connected live viewers:
6440 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6445 . On the _target system_, create a <<tracing-session,tracing session>>
6451 $ lttng create my-session --live
6455 This spawns a local relay daemon.
6457 . Start the live viewer and configure it to connect to the relay
6458 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6463 $ babeltrace --input-format=lttng-live \
6464 net://localhost/host/hostname/my-session
6471 * `hostname` with the host name of the target system.
6472 * `my-session` with the name of the tracing session to view.
6475 . Configure the tracing session as usual with the man:lttng(1)
6476 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6478 You can list the available live tracing sessions with Babeltrace:
6482 $ babeltrace --input-format=lttng-live net://localhost
6485 You can start the relay daemon on another system. In this case, you need
6486 to specify the relay daemon's URL when you create the tracing session
6487 with the opt:lttng-create(1):--set-url option. You also need to replace
6488 `localhost` in the procedure above with the host name of the system on
6489 which the relay daemon is running.
6491 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6492 command-line options.
6496 [[taking-a-snapshot]]
6497 === Take a snapshot of the current sub-buffers of a tracing session
6499 The normal behavior of LTTng is to append full sub-buffers to growing
6500 trace data files. This is ideal to keep a full history of the events
6501 that occurred on the target system, but it can
6502 represent too much data in some situations. For example, you may wish
6503 to trace your application continuously until some critical situation
6504 happens, in which case you only need the latest few recorded
6505 events to perform the desired analysis, not multi-gigabyte trace files.
6507 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6508 current sub-buffers of a given <<tracing-session,tracing session>>.
6509 LTTng can write the snapshot to the local file system or send it over
6513 .A snapshot is a copy of the current sub-buffers, which are not cleared after the operation.
6514 image::snapshot.png[]
6516 If you wish to create unmanaged, self-contained, non-overlapping
6517 trace chunk archives instead of a simple copy of the current
6518 sub-buffers, see the <<session-rotation,tracing session rotation>>
6519 feature (available since LTTng{nbsp}2.11).
6523 . Create a tracing session in _snapshot mode_:
6528 $ lttng create my-session --snapshot
6532 The <<channel-overwrite-mode-vs-discard-mode,event record loss mode>> of
6533 <<channel,channels>> created in this mode is automatically set to
6534 _overwrite_ (flight recorder mode).
6536 . Configure the tracing session as usual with the man:lttng(1)
6537 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6539 . **Optional**: When you need to take a snapshot,
6540 <<basic-tracing-session-control,stop tracing>>.
6542 You can take a snapshot when the tracers are active, but if you stop
6543 them first, you are sure that the data in the sub-buffers does not
6544 change before you actually take the snapshot.
6551 $ lttng snapshot record --name=my-first-snapshot
6555 LTTng writes the current sub-buffers of all the current tracing
6556 session's channels to trace files on the local file system. Those trace
6557 files have `my-first-snapshot` in their name.
6559 There is no difference between the format of a normal trace file and the
6560 format of a snapshot: viewers of LTTng traces also support LTTng
6563 By default, LTTng writes snapshot files to the path shown by
6564 `lttng snapshot list-output`. You can change this path or decide to send
6565 snapshots over the network using either:
6567 . An output path or URL that you specify when you
6568 <<creating-destroying-tracing-sessions,create the tracing session>>.
6569 . A snapshot output path or URL that you add using
6570 `lttng snapshot add-output`.
6571 . An output path or URL that you provide directly to the
6572 `lttng snapshot record` command.
6574 Method{nbsp}3 overrides method{nbsp}2, which overrides method 1. When
6575 you specify a URL, a relay daemon must listen on a remote system (see
6576 <<sending-trace-data-over-the-network,Send trace data over the
6581 [[session-rotation]]
6582 === Archive the current trace chunk (rotate a tracing session)
6584 The <<taking-a-snapshot,snapshot user guide>> shows how you can dump
6585 a tracing session's current sub-buffers to the file system or send them
6586 over the network. When you take a snapshot, LTTng does not clear the
6587 tracing session's ring buffers: if you take another snapshot immediately
6588 after, both snapshots could contain overlapping trace data.
6590 Inspired by https://en.wikipedia.org/wiki/Log_rotation[log rotation],
6591 _tracing session rotation_ is a feature which appends the content of the
6592 ring buffers to what's already on the file system or sent over the
6593 network since the tracing session's creation or since the last
6594 rotation, and then clears those ring buffers to avoid trace data
6597 What LTTng is about to write when performing a tracing session rotation
6598 is called the _current trace chunk_. When this current trace chunk is
6599 written to the file system or sent over the network, it becomes a _trace
6600 chunk archive_. Therefore, a tracing session rotation _archives_ the
6601 current trace chunk.
6604 .A tracing session rotation operation _archives_ the current trace chunk.
6605 image::rotation.png[]
6607 A trace chunk archive is a self-contained LTTng trace which LTTng
6608 doesn't manage anymore: you can read it, modify it, move it, or remove
6611 There are two methods to perform a tracing session rotation: immediately
6612 or with a rotation schedule.
6614 To perform an immediate tracing session rotation:
6616 . <<creating-destroying-tracing-sessions,Create a tracing session>>
6617 in _normal mode_ or _network streaming mode_
6618 (only those two creation modes support tracing session rotation):
6623 $ lttng create my-session
6627 . <<enabling-disabling-events,Create one or more event rules>>
6628 and <<basic-tracing-session-control,start tracing>>:
6633 $ lttng enable-event --kernel sched_'*'
6638 . When needed, immediately rotate the current tracing session:
6647 The cmd:lttng-rotate command prints the path to the created trace
6648 chunk archive. See man:lttng-rotate(1) to learn about the format
6649 of trace chunk archive directory names.
6651 You can perform other immediate rotations while the tracing session is
6652 active. It is guaranteed that all the trace chunk archives do not
6653 contain overlapping trace data. You can also perform an immediate
6654 rotation once you have <<basic-tracing-session-control,stopped>> the
6657 . When you are done tracing,
6658 <<creating-destroying-tracing-sessions,destroy the current tracing
6668 The tracing session destruction operation creates one last trace
6669 chunk archive from the current trace chunk.
6671 A tracing session rotation schedule is a planned rotation which LTTng
6672 performs automatically based on one of the following conditions:
6674 * A timer with a configured period times out.
6676 * The total size of the flushed part of the current trace chunk
6677 becomes greater than or equal to a configured value.
6679 To schedule a tracing session rotation, set a _rotation schedule_:
6681 . <<creating-destroying-tracing-sessions,Create a tracing session>>
6682 in _normal mode_ or _network streaming mode_
6683 (only those two creation modes support tracing session rotation):
6688 $ lttng create my-session
6692 . <<enabling-disabling-events,Create one or more event rules>>:
6697 $ lttng enable-event --kernel sched_'*'
6701 . Set a tracing session rotation schedule:
6706 $ lttng enable-rotation --timer=10s
6710 In this example, we set a rotation schedule so that LTTng performs a
6711 tracing session rotation every ten seconds.
6713 See man:lttng-enable-rotation(1) to learn more about other ways to set a
6716 . <<basic-tracing-session-control,Start tracing>>:
6725 LTTng performs tracing session rotations automatically while the tracing
6726 session is active thanks to the rotation schedule.
6728 . When you are done tracing,
6729 <<creating-destroying-tracing-sessions,destroy the current tracing
6739 The tracing session destruction operation creates one last trace chunk
6740 archive from the current trace chunk.
6742 You can use man:lttng-disable-rotation(1) to unset a tracing session
6745 NOTE: man:lttng-rotate(1) and man:lttng-enable-rotation(1) list
6746 limitations regarding those two commands.
6751 === Use the machine interface
6753 With any command of the man:lttng(1) command-line tool, you can set the
6754 opt:lttng(1):--mi option to `xml` (before the command name) to get an
6755 XML machine interface output, for example:
6759 $ lttng --mi=xml enable-event --kernel --syscall open
6762 A schema definition (XSD) is
6763 https://github.com/lttng/lttng-tools/blob/stable-2.11/src/common/mi-lttng-3.0.xsd[available]
6764 to ease the integration with external tools as much as possible.
6768 [[metadata-regenerate]]
6769 === Regenerate the metadata of an LTTng trace
6771 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
6772 data stream files and a metadata file. This metadata file contains,
6773 amongst other things, information about the offset of the clock sources
6774 used to timestamp <<event,event records>> when tracing.
6776 If, once a <<tracing-session,tracing session>> is
6777 <<basic-tracing-session-control,started>>, a major
6778 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
6779 happens, the trace's clock offset also needs to be updated. You
6780 can use the `metadata` item of the man:lttng-regenerate(1) command
6783 The main use case of this command is to allow a system to boot with
6784 an incorrect wall time and trace it with LTTng before its wall time
6785 is corrected. Once the system is known to be in a state where its
6786 wall time is correct, it can run `lttng regenerate metadata`.
6788 To regenerate the metadata of an LTTng trace:
6790 * Use the `metadata` item of the man:lttng-regenerate(1) command:
6795 $ lttng regenerate metadata
6801 `lttng regenerate metadata` has the following limitations:
6803 * Tracing session <<creating-destroying-tracing-sessions,created>>
6805 * User space <<channel,channels>>, if any, are using
6806 <<channel-buffering-schemes,per-user buffering>>.
6811 [[regenerate-statedump]]
6812 === Regenerate the state dump of a tracing session
6814 The LTTng kernel and user space tracers generate state dump
6815 <<event,event records>> when the application starts or when you
6816 <<basic-tracing-session-control,start a tracing session>>. An analysis
6817 can use the state dump event records to set an initial state before it
6818 builds the rest of the state from the following event records.
6819 http://tracecompass.org/[Trace Compass] is a notable example of an
6820 application which uses the state dump of an LTTng trace.
6822 When you <<taking-a-snapshot,take a snapshot>>, it's possible that the
6823 state dump event records are not included in the snapshot because they
6824 were recorded to a sub-buffer that has been consumed or overwritten
6827 You can use the `lttng regenerate statedump` command to emit the state
6828 dump event records again.
6830 To regenerate the state dump of the current tracing session, provided
6831 create it in snapshot mode, before you take a snapshot:
6833 . Use the `statedump` item of the man:lttng-regenerate(1) command:
6838 $ lttng regenerate statedump
6842 . <<basic-tracing-session-control,Stop the tracing session>>:
6851 . <<taking-a-snapshot,Take a snapshot>>:
6856 $ lttng snapshot record --name=my-snapshot
6860 Depending on the event throughput, you should run steps 1 and 2
6861 as closely as possible.
6863 NOTE: To record the state dump events, you need to
6864 <<enabling-disabling-events,create event rules>> which enable them.
6865 LTTng-UST state dump tracepoints start with `lttng_ust_statedump:`.
6866 LTTng-modules state dump tracepoints start with `lttng_statedump_`.
6870 [[persistent-memory-file-systems]]
6871 === Record trace data on persistent memory file systems
6873 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6874 (NVRAM) is random-access memory that retains its information when power
6875 is turned off (non-volatile). Systems with such memory can store data
6876 structures in RAM and retrieve them after a reboot, without flushing
6877 to typical _storage_.
6879 Linux supports NVRAM file systems thanks to either
6880 http://pramfs.sourceforge.net/[PRAMFS] or
6881 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6882 (requires Linux{nbsp}4.1+).
6884 This section does not describe how to operate such file systems;
6885 we assume that you have a working persistent memory file system.
6887 When you create a <<tracing-session,tracing session>>, you can specify
6888 the path of the shared memory holding the sub-buffers. If you specify a
6889 location on an NVRAM file system, then you can retrieve the latest
6890 recorded trace data when the system reboots after a crash.
6892 To record trace data on a persistent memory file system and retrieve the
6893 trace data after a system crash:
6895 . Create a tracing session with a sub-buffer shared memory path located
6896 on an NVRAM file system:
6901 $ lttng create my-session --shm-path=/path/to/shm
6905 . Configure the tracing session as usual with the man:lttng(1)
6906 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6908 . After a system crash, use the man:lttng-crash(1) command-line tool to
6909 view the trace data recorded on the NVRAM file system:
6914 $ lttng-crash /path/to/shm
6918 The binary layout of the ring buffer files is not exactly the same as
6919 the trace files layout. This is why you need to use man:lttng-crash(1)
6920 instead of your preferred trace viewer directly.
6922 To convert the ring buffer files to LTTng trace files:
6924 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
6929 $ lttng-crash --extract=/path/to/trace /path/to/shm
6935 [[notif-trigger-api]]
6936 === Get notified when a channel's buffer usage is too high or too low
6938 With LTTng's $$C/C++$$ notification and trigger API, your user
6939 application can get notified when the buffer usage of one or more
6940 <<channel,channels>> becomes too low or too high. You can use this API
6941 and enable or disable <<event,event rules>> during tracing to avoid
6942 <<channel-overwrite-mode-vs-discard-mode,discarded event records>>.
6944 .Have a user application get notified when an LTTng channel's buffer usage is too high.
6946 In this example, we create and build an application which gets notified
6947 when the buffer usage of a specific LTTng channel is higher than
6948 75{nbsp}%. We only print that it is the case in the example, but we
6949 could as well use the API of <<liblttng-ctl-lttng,`liblttng-ctl`>> to
6950 disable event rules when this happens.
6952 . Create the application's C source file:
6960 #include <lttng/domain.h>
6961 #include <lttng/action/action.h>
6962 #include <lttng/action/notify.h>
6963 #include <lttng/condition/condition.h>
6964 #include <lttng/condition/buffer-usage.h>
6965 #include <lttng/condition/evaluation.h>
6966 #include <lttng/notification/channel.h>
6967 #include <lttng/notification/notification.h>
6968 #include <lttng/trigger/trigger.h>
6969 #include <lttng/endpoint.h>
6971 int main(int argc, char *argv[])
6973 int exit_status = 0;
6974 struct lttng_notification_channel *notification_channel;
6975 struct lttng_condition *condition;
6976 struct lttng_action *action;
6977 struct lttng_trigger *trigger;
6978 const char *tracing_session_name;
6979 const char *channel_name;
6982 tracing_session_name = argv[1];
6983 channel_name = argv[2];
6986 * Create a notification channel. A notification channel
6987 * connects the user application to the LTTng session daemon.
6988 * This notification channel can be used to listen to various
6989 * types of notifications.
6991 notification_channel = lttng_notification_channel_create(
6992 lttng_session_daemon_notification_endpoint);
6995 * Create a "high buffer usage" condition. In this case, the
6996 * condition is reached when the buffer usage is greater than or
6997 * equal to 75 %. We create the condition for a specific tracing
6998 * session name, channel name, and for the user space tracing
7001 * The "low buffer usage" condition type also exists.
7003 condition = lttng_condition_buffer_usage_high_create();
7004 lttng_condition_buffer_usage_set_threshold_ratio(condition, .75);
7005 lttng_condition_buffer_usage_set_session_name(
7006 condition, tracing_session_name);
7007 lttng_condition_buffer_usage_set_channel_name(condition,
7009 lttng_condition_buffer_usage_set_domain_type(condition,
7013 * Create an action (get a notification) to take when the
7014 * condition created above is reached.
7016 action = lttng_action_notify_create();
7019 * Create a trigger. A trigger associates a condition to an
7020 * action: the action is executed when the condition is reached.
7022 trigger = lttng_trigger_create(condition, action);
7024 /* Register the trigger to LTTng. */
7025 lttng_register_trigger(trigger);
7028 * Now that we have registered a trigger, a notification will be
7029 * emitted everytime its condition is met. To receive this
7030 * notification, we must subscribe to notifications that match
7031 * the same condition.
7033 lttng_notification_channel_subscribe(notification_channel,
7037 * Notification loop. You can put this in a dedicated thread to
7038 * avoid blocking the main thread.
7041 struct lttng_notification *notification;
7042 enum lttng_notification_channel_status status;
7043 const struct lttng_evaluation *notification_evaluation;
7044 const struct lttng_condition *notification_condition;
7045 double buffer_usage;
7047 /* Receive the next notification. */
7048 status = lttng_notification_channel_get_next_notification(
7049 notification_channel, ¬ification);
7052 case LTTNG_NOTIFICATION_CHANNEL_STATUS_OK:
7054 case LTTNG_NOTIFICATION_CHANNEL_STATUS_NOTIFICATIONS_DROPPED:
7056 * The session daemon can drop notifications if
7057 * a monitoring application is not consuming the
7058 * notifications fast enough.
7061 case LTTNG_NOTIFICATION_CHANNEL_STATUS_CLOSED:
7063 * The notification channel has been closed by the
7064 * session daemon. This is typically caused by a session
7065 * daemon shutting down.
7069 /* Unhandled conditions or errors. */
7075 * A notification provides, amongst other things:
7077 * * The condition that caused this notification to be
7079 * * The condition evaluation, which provides more
7080 * specific information on the evaluation of the
7083 * The condition evaluation provides the buffer usage
7084 * value at the moment the condition was reached.
7086 notification_condition = lttng_notification_get_condition(
7088 notification_evaluation = lttng_notification_get_evaluation(
7091 /* We're subscribed to only one condition. */
7092 assert(lttng_condition_get_type(notification_condition) ==
7093 LTTNG_CONDITION_TYPE_BUFFER_USAGE_HIGH);
7096 * Get the exact sampled buffer usage from the
7097 * condition evaluation.
7099 lttng_evaluation_buffer_usage_get_usage_ratio(
7100 notification_evaluation, &buffer_usage);
7103 * At this point, instead of printing a message, we
7104 * could do something to reduce the channel's buffer
7105 * usage, like disable specific events.
7107 printf("Buffer usage is %f %% in tracing session \"%s\", "
7108 "user space channel \"%s\".\n", buffer_usage * 100,
7109 tracing_session_name, channel_name);
7110 lttng_notification_destroy(notification);
7114 lttng_action_destroy(action);
7115 lttng_condition_destroy(condition);
7116 lttng_trigger_destroy(trigger);
7117 lttng_notification_channel_destroy(notification_channel);
7123 . Build the `notif-app` application, linking it to `liblttng-ctl`:
7128 $ gcc -o notif-app notif-app.c -llttng-ctl
7132 . <<creating-destroying-tracing-sessions,Create a tracing session>>,
7133 <<enabling-disabling-events,create an event rule>> matching all the
7134 user space tracepoints, and
7135 <<basic-tracing-session-control,start tracing>>:
7140 $ lttng create my-session
7141 $ lttng enable-event --userspace --all
7146 If you create the channel manually with the man:lttng-enable-channel(1)
7147 command, you can control how frequently are the current values of the
7148 channel's properties sampled to evaluate user conditions with the
7149 opt:lttng-enable-channel(1):--monitor-timer option.
7151 . Run the `notif-app` application. This program accepts the
7152 <<tracing-session,tracing session>> name and the user space channel
7153 name as its two first arguments. The channel which LTTng automatically
7154 creates with the man:lttng-enable-event(1) command above is named
7160 $ ./notif-app my-session channel0
7164 . In another terminal, run an application with a very high event
7165 throughput so that the 75{nbsp}% buffer usage condition is reached.
7167 In the first terminal, the application should print lines like this:
7170 Buffer usage is 81.45197 % in tracing session "my-session", user space
7174 If you don't see anything, try modifying the condition in
7175 path:{notif-app.c} to a lower value (0.1, for example), rebuilding it
7176 (step{nbsp}2) and running it again (step{nbsp}4).
7183 [[lttng-modules-ref]]
7184 === noch:{LTTng-modules}
7188 [[lttng-tracepoint-enum]]
7189 ==== `LTTNG_TRACEPOINT_ENUM()` usage
7191 Use the `LTTNG_TRACEPOINT_ENUM()` macro to define an enumeration:
7195 LTTNG_TRACEPOINT_ENUM(name, TP_ENUM_VALUES(entries))
7200 * `name` with the name of the enumeration (C identifier, unique
7201 amongst all the defined enumerations).
7202 * `entries` with a list of enumeration entries.
7204 The available enumeration entry macros are:
7206 +ctf_enum_value(__name__, __value__)+::
7207 Entry named +__name__+ mapped to the integral value +__value__+.
7209 +ctf_enum_range(__name__, __begin__, __end__)+::
7210 Entry named +__name__+ mapped to the range of integral values between
7211 +__begin__+ (included) and +__end__+ (included).
7213 +ctf_enum_auto(__name__)+::
7214 Entry named +__name__+ mapped to the integral value following the
7215 last mapping's value.
7217 The last value of a `ctf_enum_value()` entry is its +__value__+
7220 The last value of a `ctf_enum_range()` entry is its +__end__+ parameter.
7222 If `ctf_enum_auto()` is the first entry in the list, its integral
7225 Use the `ctf_enum()` <<lttng-modules-tp-fields,field definition macro>>
7226 to use a defined enumeration as a tracepoint field.
7228 .Define an enumeration with `LTTNG_TRACEPOINT_ENUM()`.
7232 LTTNG_TRACEPOINT_ENUM(
7235 ctf_enum_auto("AUTO: EXPECT 0")
7236 ctf_enum_value("VALUE: 23", 23)
7237 ctf_enum_value("VALUE: 27", 27)
7238 ctf_enum_auto("AUTO: EXPECT 28")
7239 ctf_enum_range("RANGE: 101 TO 303", 101, 303)
7240 ctf_enum_auto("AUTO: EXPECT 304")
7248 [[lttng-modules-tp-fields]]
7249 ==== Tracepoint fields macros (for `TP_FIELDS()`)
7251 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
7252 tracepoint fields, which must be listed within `TP_FIELDS()` in
7253 `LTTNG_TRACEPOINT_EVENT()`, are:
7255 [role="func-desc growable",cols="asciidoc,asciidoc"]
7256 .Available macros to define LTTng-modules tracepoint fields
7258 |Macro |Description and parameters
7261 +ctf_integer(__t__, __n__, __e__)+
7263 +ctf_integer_nowrite(__t__, __n__, __e__)+
7265 +ctf_user_integer(__t__, __n__, __e__)+
7267 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
7269 Standard integer, displayed in base{nbsp}10.
7272 Integer C type (`int`, `long`, `size_t`, ...).
7278 Argument expression.
7281 +ctf_integer_hex(__t__, __n__, __e__)+
7283 +ctf_user_integer_hex(__t__, __n__, __e__)+
7285 Standard integer, displayed in base{nbsp}16.
7294 Argument expression.
7296 |+ctf_integer_oct(__t__, __n__, __e__)+
7298 Standard integer, displayed in base{nbsp}8.
7307 Argument expression.
7310 +ctf_integer_network(__t__, __n__, __e__)+
7312 +ctf_user_integer_network(__t__, __n__, __e__)+
7314 Integer in network byte order (big-endian), displayed in base{nbsp}10.
7323 Argument expression.
7326 +ctf_integer_network_hex(__t__, __n__, __e__)+
7328 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
7330 Integer in network byte order, displayed in base{nbsp}16.
7339 Argument expression.
7342 +ctf_enum(__N__, __t__, __n__, __e__)+
7344 +ctf_enum_nowrite(__N__, __t__, __n__, __e__)+
7346 +ctf_user_enum(__N__, __t__, __n__, __e__)+
7348 +ctf_user_enum_nowrite(__N__, __t__, __n__, __e__)+
7353 Name of a <<lttng-tracepoint-enum,previously defined enumeration>>.
7356 Integer C type (`int`, `long`, `size_t`, ...).
7362 Argument expression.
7365 +ctf_string(__n__, __e__)+
7367 +ctf_string_nowrite(__n__, __e__)+
7369 +ctf_user_string(__n__, __e__)+
7371 +ctf_user_string_nowrite(__n__, __e__)+
7373 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
7379 Argument expression.
7382 +ctf_array(__t__, __n__, __e__, __s__)+
7384 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
7386 +ctf_user_array(__t__, __n__, __e__, __s__)+
7388 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
7390 Statically-sized array of integers.
7393 Array element C type.
7399 Argument expression.
7405 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
7407 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7409 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
7411 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7413 Statically-sized array of bits.
7415 The type of +__e__+ must be an integer type. +__s__+ is the number
7416 of elements of such type in +__e__+, not the number of bits.
7419 Array element C type.
7425 Argument expression.
7431 +ctf_array_text(__t__, __n__, __e__, __s__)+
7433 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
7435 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
7437 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
7439 Statically-sized array, printed as text.
7441 The string does not need to be null-terminated.
7444 Array element C type (always `char`).
7450 Argument expression.
7456 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
7458 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7460 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
7462 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7464 Dynamically-sized array of integers.
7466 The type of +__E__+ must be unsigned.
7469 Array element C type.
7475 Argument expression.
7478 Length expression C type.
7484 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7486 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7488 Dynamically-sized array of integers, displayed in base{nbsp}16.
7490 The type of +__E__+ must be unsigned.
7493 Array element C type.
7499 Argument expression.
7502 Length expression C type.
7507 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
7509 Dynamically-sized array of integers in network byte order (big-endian),
7510 displayed in base{nbsp}10.
7512 The type of +__E__+ must be unsigned.
7515 Array element C type.
7521 Argument expression.
7524 Length expression C type.
7530 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7532 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7534 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7536 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7538 Dynamically-sized array of bits.
7540 The type of +__e__+ must be an integer type. +__s__+ is the number
7541 of elements of such type in +__e__+, not the number of bits.
7543 The type of +__E__+ must be unsigned.
7546 Array element C type.
7552 Argument expression.
7555 Length expression C type.
7561 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7563 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7565 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7567 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7569 Dynamically-sized array, displayed as text.
7571 The string does not need to be null-terminated.
7573 The type of +__E__+ must be unsigned.
7575 The behaviour is undefined if +__e__+ is `NULL`.
7578 Sequence element C type (always `char`).
7584 Argument expression.
7587 Length expression C type.
7593 Use the `_user` versions when the argument expression, `e`, is
7594 a user space address. In the cases of `ctf_user_integer*()` and
7595 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7598 The `_nowrite` versions omit themselves from the session trace, but are
7599 otherwise identical. This means the `_nowrite` fields won't be written
7600 in the recorded trace. Their primary purpose is to make some
7601 of the event context available to the
7602 <<enabling-disabling-events,event filters>> without having to
7603 commit the data to sub-buffers.
7609 Terms related to LTTng and to tracing in general:
7612 The http://diamon.org/babeltrace[Babeltrace] project, which includes:
7614 * The cmd:babeltrace (Babeltrace{nbsp}1) or cmd:babeltrace2
7615 (Babeltrace{nbsp}2) command.
7616 * Libraries with a C{nbsp}API.
7617 * Python{nbsp}3 bindings.
7618 * Plugins (Babeltrace{nbsp}2).
7620 [[def-buffering-scheme]]<<channel-buffering-schemes,buffering scheme>>::
7621 A layout of <<def-sub-buffer,sub-buffers>> applied to a given channel.
7623 [[def-channel]]<<channel,channel>>::
7624 An entity which is responsible for a set of
7625 <<def-ring-buffer,ring buffers>>.
7627 <<def-event-rule,Event rules>> are always attached to a specific
7631 A source of time for a <<def-tracer,tracer>>.
7633 [[def-consumer-daemon]]<<lttng-consumerd,consumer daemon>>::
7634 A process which is responsible for consuming the full
7635 <<def-sub-buffer,sub-buffers>> and write them to a file system or
7636 send them over the network.
7638 [[def-current-trace-chunk]]current trace chunk::
7639 A <<def-trace-chunk,trace chunk>> which includes the current content
7640 of all the <<def-tracing-session-rotation,tracing session>>'s
7641 <<def-sub-buffer,sub-buffers>> and the stream files produced since the
7642 latest event amongst:
7644 * The creation of the <<def-tracing-session,tracing session>>.
7645 * The last tracing session rotation, if any.
7647 <<channel-overwrite-mode-vs-discard-mode,discard mode>>::
7648 The <<def-event-record-loss-mode,event record loss mode>> in which
7649 the <<def-tracer,tracer>> _discards_ new event records when there's no
7650 <<def-sub-buffer,sub-buffer>> space left to store them.
7652 [[def-event]]event::
7653 The consequence of the execution of an
7654 <<def-instrumentation-point,instrumentation point>>, like a
7655 <<def-tracepoint,tracepoint>> that you manually place in some source
7656 code, or a Linux kernel kprobe.
7658 An event is said to _occur_ at a specific time. <<def-lttng,LTTng>> can
7659 take various actions upon the occurrence of an event, like record the
7660 event's payload to a <<def-sub-buffer,sub-buffer>>.
7662 [[def-event-name]]event name::
7663 The name of an <<def-event,event>>, which is also the name of the
7664 <<def-event-record,event record>>.
7666 This is also called the _instrumentation point name_.
7668 [[def-event-record]]event record::
7669 A record, in a <<def-trace,trace>>, of the payload of an
7670 <<def-event,event>> which occured.
7672 [[def-event-record-loss-mode]]<<channel-overwrite-mode-vs-discard-mode,event record loss mode>>::
7673 The mechanism by which event records of a given
7674 <<def-channel,channel>> are lost (not recorded) when there is no
7675 <<def-sub-buffer,sub-buffer>> space left to store them.
7677 [[def-event-rule]]<<event,event rule>>::
7678 Set of conditions which must be satisfied for one or more occuring
7679 <<def-event,events>> to be recorded.
7681 `java.util.logging`::
7683 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
7685 <<instrumenting,instrumentation>>::
7686 The use of <<def-lttng,LTTng>> probes to make a piece of software
7689 [[def-instrumentation-point]]instrumentation point::
7690 A point in the execution path of a piece of software that, when
7691 reached by this execution, can emit an <<def-event,event>>.
7693 instrumentation point name::
7694 See _<<def-event-name,event name>>_.
7697 A http://logging.apache.org/log4j/1.2/[logging library] for Java
7698 developed by the Apache Software Foundation.
7701 Level of severity of a log statement or user space
7702 <<def-instrumentation-point,instrumentation point>>.
7704 [[def-lttng]]LTTng::
7705 The _Linux Trace Toolkit: next generation_ project.
7707 <<lttng-cli,cmd:lttng>>::
7708 A command-line tool provided by the <<def-lttng-tools,LTTng-tools>>
7709 project which you can use to send and receive control messages to and
7710 from a <<def-session-daemon,session daemon>>.
7713 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
7714 which is a set of analyzing programs that you can use to obtain a
7715 higher level view of an <<def-lttng,LTTng>> <<def-trace,trace>>.
7717 cmd:lttng-consumerd::
7718 The name of the <<def-consumer-daemon,consumer daemon>> program.
7721 A utility provided by the <<def-lttng-tools,LTTng-tools>> project
7722 which can convert <<def-ring-buffer,ring buffer>> files (usually
7723 <<persistent-memory-file-systems,saved on a persistent memory file
7724 system>>) to <<def-trace,trace>> files.
7726 See man:lttng-crash(1).
7728 LTTng Documentation::
7731 <<lttng-live,LTTng live>>::
7732 A communication protocol between the <<lttng-relayd,relay daemon>> and
7733 live viewers which makes it possible to see <<def-event-record,event
7734 records>> "live", as they are received by the
7735 <<def-relay-daemon,relay daemon>>.
7737 <<lttng-modules,LTTng-modules>>::
7738 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
7739 which contains the Linux kernel modules to make the Linux kernel
7740 <<def-instrumentation-point,instrumentation points>> available for
7741 <<def-lttng,LTTng>> tracing.
7744 The name of the <<def-relay-daemon,relay daemon>> program.
7746 cmd:lttng-sessiond::
7747 The name of the <<def-session-daemon,session daemon>> program.
7749 [[def-lttng-tools]]LTTng-tools::
7750 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
7751 contains the various programs and libraries used to
7752 <<controlling-tracing,control tracing>>.
7754 [[def-lttng-ust]]<<lttng-ust,LTTng-UST>>::
7755 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
7756 contains libraries to instrument
7757 <<def-user-application,user applications>>.
7759 <<lttng-ust-agents,LTTng-UST Java agent>>::
7760 A Java package provided by the <<def-lttng-ust,LTTng-UST>> project to
7761 allow the LTTng instrumentation of `java.util.logging` and Apache
7762 log4j{nbsp}1.2 logging statements.
7764 <<lttng-ust-agents,LTTng-UST Python agent>>::
7765 A Python package provided by the <<def-lttng-ust,LTTng-UST>> project
7766 to allow the <<def-lttng,LTTng>> instrumentation of Python logging
7769 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
7770 The <<def-event-record-loss-mode,event record loss mode>> in which new
7771 <<def-event-record,event records>> _overwrite_ older event records
7772 when there's no <<def-sub-buffer,sub-buffer>> space left to store
7775 <<channel-buffering-schemes,per-process buffering>>::
7776 A <<def-buffering-scheme,buffering scheme>> in which each instrumented
7777 process has its own <<def-sub-buffer,sub-buffers>> for a given user
7778 space <<def-channel,channel>>.
7780 <<channel-buffering-schemes,per-user buffering>>::
7781 A <<def-buffering-scheme,buffering scheme>> in which all the processes
7782 of a Unix user share the same <<def-sub-buffer,sub-buffers>> for a
7783 given user space <<def-channel,channel>>.
7785 [[def-relay-daemon]]<<lttng-relayd,relay daemon>>::
7786 A process which is responsible for receiving the <<def-trace,trace>>
7787 data which a distant <<def-consumer-daemon,consumer daemon>> sends.
7789 [[def-ring-buffer]]ring buffer::
7790 A set of <<def-sub-buffer,sub-buffers>>.
7793 See _<<def-tracing-session-rotation,tracing session rotation>>_.
7795 [[def-session-daemon]]<<lttng-sessiond,session daemon>>::
7796 A process which receives control commands from you and orchestrates
7797 the <<def-tracer,tracers>> and various <<def-lttng,LTTng>> daemons.
7799 <<taking-a-snapshot,snapshot>>::
7800 A copy of the current data of all the <<def-sub-buffer,sub-buffers>>
7801 of a given <<def-tracing-session,tracing session>>, saved as
7802 <<def-trace,trace>> files.
7804 [[def-sub-buffer]]sub-buffer::
7805 One part of an <<def-lttng,LTTng>> <<def-ring-buffer,ring buffer>>
7806 which contains <<def-event-record,event records>>.
7809 The time information attached to an <<def-event,event>> when it is
7812 [[def-trace]]trace (_noun_)::
7815 * One http://diamon.org/ctf/[CTF] metadata stream file.
7816 * One or more CTF data stream files which are the concatenations of one
7817 or more flushed <<def-sub-buffer,sub-buffers>>.
7819 [[def-trace-verb]]trace (_verb_)::
7820 The action of recording the <<def-event,events>> emitted by an
7821 application or by a system, or to initiate such recording by
7822 controlling a <<def-tracer,tracer>>.
7824 [[def-trace-chunk]]trace chunk::
7825 A self-contained <<def-trace,trace>> which is part of a
7826 <<def-tracing-session,tracing session>>. Each
7827 <<def-tracing-session-rotation, tracing session rotation>> produces a
7828 <<def-trace-chunk-archive,trace chunk archive>>.
7830 [[def-trace-chunk-archive]]trace chunk archive::
7831 The result of a <<def-tracing-session-rotation, tracing session rotation>>.
7833 <<def-lttng,LTTng>> does not manage any trace chunk archive, even if its
7834 containing <<def-tracing-session,tracing session>> is still active: you
7835 are free to read it, modify it, move it, or remove it.
7838 The http://tracecompass.org[Trace Compass] project and application.
7840 [[def-tracepoint]]tracepoint::
7841 An instrumentation point using the tracepoint mechanism of the Linux
7842 kernel or of <<def-lttng-ust,LTTng-UST>>.
7844 tracepoint definition::
7845 The definition of a single <<def-tracepoint,tracepoint>>.
7848 The name of a <<def-tracepoint,tracepoint>>.
7850 [[def-tracepoint-provider]]tracepoint provider::
7851 A set of functions providing <<def-tracepoint,tracepoints>> to an
7852 instrumented <<def-user-application,user application>>.
7854 Not to be confused with a <<def-tracepoint-provider-package,tracepoint
7855 provider package>>: many tracepoint providers can exist within a
7856 tracepoint provider package.
7858 [[def-tracepoint-provider-package]]tracepoint provider package::
7859 One or more <<def-tracepoint-provider,tracepoint providers>> compiled
7860 as an https://en.wikipedia.org/wiki/Object_file[object file] or as a
7861 link:https://en.wikipedia.org/wiki/Library_(computing)#Shared_libraries[shared
7864 [[def-tracer]]tracer::
7865 A software which records emitted <<def-event,events>>.
7867 <<domain,tracing domain>>::
7868 A namespace for <<def-event,event>> sources.
7870 <<tracing-group,tracing group>>::
7871 The Unix group in which a Unix user can be to be allowed to
7872 <<def-trace-verb,trace>> the Linux kernel.
7874 [[def-tracing-session]]<<tracing-session,tracing session>>::
7875 A stateful dialogue between you and a <<lttng-sessiond,session daemon>>.
7877 [[def-tracing-session-rotation]]<<session-rotation,tracing session rotation>>::
7878 The action of archiving the
7879 <<def-current-trace-chunk,current trace chunk>> of a
7880 <<def-tracing-session,tracing session>>.
7882 [[def-user-application]]user application::
7883 An application running in user space, as opposed to a Linux kernel
7884 module, for example.