1 The LTTng Documentation
2 =======================
3 Philippe Proulx <pproulx@efficios.com>
4 v2.10, 19 November 2018
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 {revision}?
77 LTTng{nbsp}{revision} bears the name _KeKriek_. From
78 http://brasseriedunham.com/[Brasserie Dunham], the _**KeKriek**_ is a
79 sour mashed golden wheat ale fermented with local sour cherries from
80 Tougas orchards. Fresh sweet cherry notes with some tartness, lively
81 carbonation with a dry finish.
83 New features and changes in LTTng{nbsp}{revision}:
85 * **Tracing control**:
86 ** You can put more than one wildcard special character (`*`), and not
87 only at the end, when you <<enabling-disabling-events,create an event
88 rule>>, in both the instrumentation point name and the literal
90 link:/man/1/lttng-enable-event/v{revision}/#doc-filter-syntax[filter expressions]:
95 # lttng enable-event --kernel 'x86_*_local_timer_*' \
96 --filter='name == "*a*b*c*d*e" && count >= 23'
103 $ lttng enable-event --userspace '*_my_org:*msg*'
107 ** New trigger and notification API for
108 <<liblttng-ctl-lttng,`liblttng-ctl`>>. This new subsystem allows you
109 to register triggers which emit a notification when a given
110 condition is satisfied. As of LTTng{nbsp}{revision}, only
111 <<channel,channel>> buffer usage conditions are available.
112 Documentation is available in the
113 https://github.com/lttng/lttng-tools/tree/stable-{revision}/include/lttng[`liblttng-ctl`
115 <<notif-trigger-api,Get notified when a channel's buffer usage is too
118 ** You can now embed the whole textual LTTng-tools man pages into the
119 executables at build time with the `--enable-embedded-help`
120 configuration option. Thanks to this option, you don't need the
121 http://www.methods.co.nz/asciidoc/[AsciiDoc] and
122 https://directory.fsf.org/wiki/Xmlto[xmlto] tools at build time, and
123 a manual pager at run time, to get access to this documentation.
125 * **User space tracing**:
126 ** New blocking mode: an LTTng-UST tracepoint can now block until
127 <<channel,sub-buffer>> space is available instead of discarding event
128 records in <<channel-overwrite-mode-vs-discard-mode,discard mode>>.
129 With this feature, you can be sure that no event records are
130 discarded during your application's execution at the expense of
133 For example, the following command lines create a user space tracing
134 channel with an infinite blocking timeout and run an application
135 instrumented with LTTng-UST which is explicitly allowed to block:
141 $ lttng enable-channel --userspace --blocking-timeout=inf blocking-channel
142 $ lttng enable-event --userspace --channel=blocking-channel --all
144 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
148 See the complete <<blocking-timeout-example,blocking timeout example>>.
150 * **Linux kernel tracing**:
151 ** Linux 4.10, 4.11, and 4.12 support.
152 ** The thread state dump events recorded by LTTng-modules now contain
153 the task's CPU identifier. This improves the precision of the
154 scheduler model for analyses.
155 ** Extended man:socketpair(2) system call tracing data.
161 What is LTTng? As its name suggests, the _Linux Trace Toolkit: next
162 generation_ is a modern toolkit for tracing Linux systems and
163 applications. So your first question might be:
170 As the history of software engineering progressed and led to what
171 we now take for granted--complex, numerous and
172 interdependent software applications running in parallel on
173 sophisticated operating systems like Linux--the authors of such
174 components, software developers, began feeling a natural
175 urge to have tools that would ensure the robustness and good performance
176 of their masterpieces.
178 One major achievement in this field is, inarguably, the
179 https://www.gnu.org/software/gdb/[GNU debugger (GDB)],
180 an essential tool for developers to find and fix bugs. But even the best
181 debugger won't help make your software run faster, and nowadays, faster
182 software means either more work done by the same hardware, or cheaper
183 hardware for the same work.
185 A _profiler_ is often the tool of choice to identify performance
186 bottlenecks. Profiling is suitable to identify _where_ performance is
187 lost in a given software. The profiler outputs a profile, a statistical
188 summary of observed events, which you may use to discover which
189 functions took the most time to execute. However, a profiler won't
190 report _why_ some identified functions are the bottleneck. Bottlenecks
191 might only occur when specific conditions are met, conditions that are
192 sometimes impossible to capture by a statistical profiler, or impossible
193 to reproduce with an application altered by the overhead of an
194 event-based profiler. For a thorough investigation of software
195 performance issues, a history of execution is essential, with the
196 recorded values of variables and context fields you choose, and
197 with as little influence as possible on the instrumented software. This
198 is where tracing comes in handy.
200 _Tracing_ is a technique used to understand what goes on in a running
201 software system. The software used for tracing is called a _tracer_,
202 which is conceptually similar to a tape recorder. When recording,
203 specific instrumentation points placed in the software source code
204 generate events that are saved on a giant tape: a _trace_ file. You
205 can trace user applications and the operating system at the same time,
206 opening the possibility of resolving a wide range of problems that would
207 otherwise be extremely challenging.
209 Tracing is often compared to _logging_. However, tracers and loggers are
210 two different tools, serving two different purposes. Tracers are
211 designed to record much lower-level events that occur much more
212 frequently than log messages, often in the range of thousands per
213 second, with very little execution overhead. Logging is more appropriate
214 for a very high-level analysis of less frequent events: user accesses,
215 exceptional conditions (errors and warnings, for example), database
216 transactions, instant messaging communications, and such. Simply put,
217 logging is one of the many use cases that can be satisfied with tracing.
219 The list of recorded events inside a trace file can be read manually
220 like a log file for the maximum level of detail, but it is generally
221 much more interesting to perform application-specific analyses to
222 produce reduced statistics and graphs that are useful to resolve a
223 given problem. Trace viewers and analyzers are specialized tools
226 In the end, this is what LTTng is: a powerful, open source set of
227 tools to trace the Linux kernel and user applications at the same time.
228 LTTng is composed of several components actively maintained and
229 developed by its link:/community/#where[community].
232 [[lttng-alternatives]]
233 === Alternatives to noch:{LTTng}
235 Excluding proprietary solutions, a few competing software tracers
238 * https://github.com/dtrace4linux/linux[dtrace4linux] is a port of
239 Sun Microsystems's DTrace to Linux. The cmd:dtrace tool interprets
240 user scripts and is responsible for loading code into the
241 Linux kernel for further execution and collecting the outputted data.
242 * https://en.wikipedia.org/wiki/Berkeley_Packet_Filter[eBPF] is a
243 subsystem in the Linux kernel in which a virtual machine can execute
244 programs passed from the user space to the kernel. You can attach
245 such programs to tracepoints and KProbes thanks to a system call, and
246 they can output data to the user space when executed thanks to
247 different mechanisms (pipe, VM register values, and eBPF maps, to name
249 * https://www.kernel.org/doc/Documentation/trace/ftrace.txt[ftrace]
250 is the de facto function tracer of the Linux kernel. Its user
251 interface is a set of special files in sysfs.
252 * https://perf.wiki.kernel.org/[perf] is
253 a performance analyzing tool for Linux which supports hardware
254 performance counters, tracepoints, as well as other counters and
255 types of probes. perf's controlling utility is the cmd:perf command
257 * http://linux.die.net/man/1/strace[strace]
258 is a command-line utility which records system calls made by a
259 user process, as well as signal deliveries and changes of process
260 state. strace makes use of https://en.wikipedia.org/wiki/Ptrace[ptrace]
261 to fulfill its function.
262 * http://www.sysdig.org/[sysdig], like SystemTap, uses scripts to
263 analyze Linux kernel events. You write scripts, or _chisels_ in
264 sysdig's jargon, in Lua and sysdig executes them while the system is
265 being traced or afterwards. sysdig's interface is the cmd:sysdig
266 command-line tool as well as the curses-based cmd:csysdig tool.
267 * https://sourceware.org/systemtap/[SystemTap] is a Linux kernel and
268 user space tracer which uses custom user scripts to produce plain text
269 traces. SystemTap converts the scripts to the C language, and then
270 compiles them as Linux kernel modules which are loaded to produce
271 trace data. SystemTap's primary user interface is the cmd:stap
274 The main distinctive features of LTTng is that it produces correlated
275 kernel and user space traces, as well as doing so with the lowest
276 overhead amongst other solutions. It produces trace files in the
277 http://diamon.org/ctf[CTF] format, a file format optimized
278 for the production and analyses of multi-gigabyte data.
280 LTTng is the result of more than 10 years of active open source
281 development by a community of passionate developers.
282 LTTng{nbsp}{revision} is currently available on major desktop and server
285 The main interface for tracing control is a single command-line tool
286 named cmd:lttng. The latter can create several tracing sessions, enable
287 and disable events on the fly, filter events efficiently with custom
288 user expressions, start and stop tracing, and much more. LTTng can
289 record the traces on the file system or send them over the network, and
290 keep them totally or partially. You can view the traces once tracing
291 becomes inactive or in real-time.
293 <<installing-lttng,Install LTTng now>> and
294 <<getting-started,start tracing>>!
300 **LTTng** is a set of software <<plumbing,components>> which interact to
301 <<instrumenting,instrument>> the Linux kernel and user applications, and
302 to <<controlling-tracing,control tracing>> (start and stop
303 tracing, enable and disable event rules, and the rest). Those
304 components are bundled into the following packages:
306 * **LTTng-tools**: Libraries and command-line interface to
308 * **LTTng-modules**: Linux kernel modules to instrument and
310 * **LTTng-UST**: Libraries and Java/Python packages to instrument and
311 trace user applications.
313 Most distributions mark the LTTng-modules and LTTng-UST packages as
314 optional when installing LTTng-tools (which is always required). In the
315 following sections, we always provide the steps to install all three,
318 * You only need to install LTTng-modules if you intend to trace the
320 * You only need to install LTTng-UST if you intend to trace user
324 .Availability of LTTng{nbsp}{revision} for major Linux distributions as of 12 November 2018.
326 |Distribution |Available in releases |Alternatives
328 |https://www.ubuntu.com/[Ubuntu]
329 |<<ubuntu,Ubuntu{nbsp}18.04 _Bionic Beaver_>>.
331 Ubuntu{nbsp}14.04 _Trusty Tahr_, Ubuntu{nbsp}16.04 _Xenial Xerus_,
332 and Ubuntu{nbsp}18.04 _Bionic Beaver_:
333 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
334 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
336 |https://getfedora.org/[Fedora]
337 |xref:fedora[Fedora{nbsp}27, Fedora{nbsp}28, and Fedora{nbsp}29].
338 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
340 |https://www.debian.org/[Debian]
341 |<<debian,Debian "buster" (testing)>>.
342 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
344 |https://www.archlinux.org/[Arch Linux]
345 |<<arch-linux,Current Arch Linux build>>.
346 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
348 |https://alpinelinux.org/[Alpine Linux]
349 |<<alpine-linux,Alpine Linux{nbsp}3.7 and Alpine Linux{nbsp}3.8>>.
350 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
352 |https://www.opensuse.org/[openSUSE]
353 |<<opensuse,openSUSE Leap{nbsp}15.0>>.
354 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
356 |https://www.redhat.com/[RHEL] and https://www.suse.com/[SLES]
357 |See http://packages.efficios.com/[EfficiOS Enterprise Packages].
360 |https://buildroot.org/[Buildroot]
361 |xref:buildroot[Buildroot{nbsp}2018.02, Buildroot{nbsp}2018.05,
362 Buildroot{nbsp}2018.08, and Buildroot{nbsp}2018.11].
363 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
365 |http://www.openembedded.org/wiki/Main_Page[OpenEmbedded] and
366 https://www.yoctoproject.org/[Yocto]
368 |<<building-from-source,Build LTTng{nbsp}{revision} from source>>.
373 === [[ubuntu-official-repositories]]Ubuntu
375 LTTng{nbsp}{revision} is available on Ubuntu{nbsp}18.04 _Bionic Beaver_.
376 For previous supported releases of Ubuntu,
377 <<ubuntu-ppa,use the LTTng Stable{nbsp}{revision} PPA>>.
379 To install LTTng{nbsp}{revision} on Ubuntu{nbsp}18.04 _Bionic Beaver_:
381 . Install the main LTTng{nbsp}{revision} packages:
386 # apt-get install lttng-tools
387 # apt-get install lttng-modules-dkms
388 # apt-get install liblttng-ust-dev
392 . **If you need to instrument and trace
393 <<java-application,Java applications>>**, install the LTTng-UST
399 # apt-get install liblttng-ust-agent-java
403 . **If you need to instrument and trace
404 <<python-application,Python{nbsp}3 applications>>**, install the
405 LTTng-UST Python agent:
410 # apt-get install python3-lttngust
416 ==== noch:{LTTng} Stable {revision} PPA
418 The https://launchpad.net/~lttng/+archive/ubuntu/stable-{revision}[LTTng
419 Stable{nbsp}{revision} PPA] offers the latest stable
420 LTTng{nbsp}{revision} packages for:
422 * Ubuntu{nbsp}14.04 _Trusty Tahr_
423 * Ubuntu{nbsp}16.04 _Xenial Xerus_
424 * Ubuntu{nbsp}18.04 _Bionic Beaver_
426 To install LTTng{nbsp}{revision} from the LTTng Stable{nbsp}{revision} PPA:
428 . Add the LTTng Stable{nbsp}{revision} PPA repository and update the
434 # apt-add-repository ppa:lttng/stable-2.10
439 . Install the main LTTng{nbsp}{revision} packages:
444 # apt-get install lttng-tools
445 # apt-get install lttng-modules-dkms
446 # apt-get install liblttng-ust-dev
450 . **If you need to instrument and trace
451 <<java-application,Java applications>>**, install the LTTng-UST
457 # apt-get install liblttng-ust-agent-java
461 . **If you need to instrument and trace
462 <<python-application,Python{nbsp}3 applications>>**, install the
463 LTTng-UST Python agent:
468 # apt-get install python3-lttngust
476 To install LTTng{nbsp}{revision} on Fedora{nbsp}27, Fedora{nbsp}28,
479 . Install the LTTng-tools{nbsp}{revision} and LTTng-UST{nbsp}{revision}
485 # yum install lttng-tools
486 # yum install lttng-ust
490 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
496 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
497 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
498 cd lttng-modules-2.10.* &&
500 sudo make modules_install &&
506 .Java and Python application instrumentation and tracing
508 If you need to instrument and trace <<java-application,Java
509 applications>> on Fedora, you need to build and install
510 LTTng-UST{nbsp}{revision} <<building-from-source,from source>> and pass
511 the `--enable-java-agent-jul`, `--enable-java-agent-log4j`, or
512 `--enable-java-agent-all` options to the `configure` script, depending
513 on which Java logging framework you use.
515 If you need to instrument and trace <<python-application,Python
516 applications>> on Fedora, you need to build and install
517 LTTng-UST{nbsp}{revision} from source and pass the
518 `--enable-python-agent` option to the `configure` script.
525 To install LTTng{nbsp}{revision} on Debian "buster" (testing):
527 . Install the main LTTng{nbsp}{revision} packages:
532 # apt-get install lttng-modules-dkms
533 # apt-get install liblttng-ust-dev
534 # apt-get install lttng-tools
538 . **If you need to instrument and trace <<java-application,Java
539 applications>>**, install the LTTng-UST Java agent:
544 # apt-get install liblttng-ust-agent-java
548 . **If you need to instrument and trace <<python-application,Python
549 applications>>**, install the LTTng-UST Python agent:
554 # apt-get install python3-lttngust
562 LTTng-UST{nbsp}{revision} is available in Arch Linux's _Community_
563 repository, while LTTng-tools{nbsp}{revision} and
564 LTTng-modules{nbsp}{revision} are available in the
565 https://aur.archlinux.org/[AUR].
567 To install LTTng{nbsp}{revision} on Arch Linux, using
568 https://github.com/rmarquis/pacaur[pacaur] for the AUR packages:
570 . Install the main LTTng{nbsp}{revision} packages:
575 # pacman -Sy lttng-ust
576 $ pikaur -Sy lttng-tools
577 $ pikaur -Sy lttng-modules
581 . **If you need to instrument and trace <<python-application,Python
582 applications>>**, install the LTTng-UST Python agent:
587 # pacman -Sy python-lttngust
588 # pacman -Sy python2-lttngust
596 To install LTTng-tools{nbsp}{revision} and LTTng-UST{nbsp}{revision} on
597 Alpine Linux{nbsp}3.7 or Alpine Linux{nbsp}3.8:
599 . Add the LTTng packages:
604 # apk add lttng-tools
605 # apk add lttng-ust-dev
609 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
615 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
616 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
617 cd lttng-modules-2.10.* &&
619 sudo make modules_install &&
628 To install LTTng{nbsp}{revision} on openSUSE Leap{nbsp}15.0:
630 * Install the main LTTng{nbsp}{revision} packages:
635 sudo zypper install lttng-tools
636 sudo zypper install lttng-modules
637 sudo zypper install lttng-ust-devel
642 .Java and Python application instrumentation and tracing
644 If you need to instrument and trace <<java-application,Java
645 applications>> on openSUSE, you need to build and install
646 LTTng-UST{nbsp}{revision} <<building-from-source,from source>> and pass
647 the `--enable-java-agent-jul`, `--enable-java-agent-log4j`, or
648 `--enable-java-agent-all` options to the `configure` script, depending
649 on which Java logging framework you use.
651 If you need to instrument and trace <<python-application,Python
652 applications>> on openSUSE, you need to build and install
653 LTTng-UST{nbsp}{revision} from source and pass the
654 `--enable-python-agent` option to the `configure` script.
658 [[enterprise-distributions]]
659 === RHEL, SUSE, and other enterprise distributions
661 To install LTTng on enterprise Linux distributions, such as Red Hat
662 Enterprise Linux (RHEL) and SUSE Linux Enterprise Server (SUSE), please
663 see http://packages.efficios.com/[EfficiOS Enterprise Packages].
669 To install LTTng{nbsp}{revision} on Buildroot{nbsp}2018.02,
670 Buildroot{nbsp}2018.05, Buildroot{nbsp}2018.08, or
671 Buildroot{nbsp}2018.11:
673 . Launch the Buildroot configuration tool:
682 . In **Kernel**, check **Linux kernel**.
683 . In **Toolchain**, check **Enable WCHAR support**.
684 . In **Target packages**{nbsp}→ **Debugging, profiling and benchmark**,
685 check **lttng-modules** and **lttng-tools**.
686 . In **Target packages**{nbsp}→ **Libraries**{nbsp}→
687 **Other**, check **lttng-libust**.
690 [[building-from-source]]
691 === Build from source
693 To build and install LTTng{nbsp}{revision} from source:
695 . Using your distribution's package manager, or from source, install
696 the following dependencies of LTTng-tools and LTTng-UST:
699 * https://sourceforge.net/projects/libuuid/[libuuid]
700 * http://directory.fsf.org/wiki/Popt[popt]
701 * http://liburcu.org/[Userspace RCU]
702 * http://www.xmlsoft.org/[libxml2]
705 . Download, build, and install the latest LTTng-modules{nbsp}{revision}:
711 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
712 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
713 cd lttng-modules-2.10.* &&
715 sudo make modules_install &&
720 . Download, build, and install the latest LTTng-UST{nbsp}{revision}:
726 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.10.tar.bz2 &&
727 tar -xf lttng-ust-latest-2.10.tar.bz2 &&
728 cd lttng-ust-2.10.* &&
738 .Java and Python application tracing
740 If you need to instrument and trace <<java-application,Java
741 applications>>, pass the `--enable-java-agent-jul`,
742 `--enable-java-agent-log4j`, or `--enable-java-agent-all` options to the
743 `configure` script, depending on which Java logging framework you use.
745 If you need to instrument and trace <<python-application,Python
746 applications>>, pass the `--enable-python-agent` option to the
747 `configure` script. You can set the `PYTHON` environment variable to the
748 path to the Python interpreter for which to install the LTTng-UST Python
756 By default, LTTng-UST libraries are installed to
757 dir:{/usr/local/lib}, which is the de facto directory in which to
758 keep self-compiled and third-party libraries.
760 When <<building-tracepoint-providers-and-user-application,linking an
761 instrumented user application with `liblttng-ust`>>:
763 * Append `/usr/local/lib` to the env:LD_LIBRARY_PATH environment
765 * Pass the `-L/usr/local/lib` and `-Wl,-rpath,/usr/local/lib` options to
766 man:gcc(1), man:g++(1), or man:clang(1).
770 . Download, build, and install the latest LTTng-tools{nbsp}{revision}:
776 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
777 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
778 cd lttng-tools-2.10.* &&
786 TIP: The https://github.com/eepp/vlttng[vlttng tool] can do all the
787 previous steps automatically for a given version of LTTng and confine
788 the installed files in a specific directory. This can be useful to test
789 LTTng without installing it on your system.
795 This is a short guide to get started quickly with LTTng kernel and user
798 Before you follow this guide, make sure to <<installing-lttng,install>>
801 This tutorial walks you through the steps to:
803 . <<tracing-the-linux-kernel,Trace the Linux kernel>>.
804 . <<tracing-your-own-user-application,Trace a user application>> written
806 . <<viewing-and-analyzing-your-traces,View and analyze the
810 [[tracing-the-linux-kernel]]
811 === Trace the Linux kernel
813 The following command lines start with the `#` prompt because you need
814 root privileges to trace the Linux kernel. You can also trace the kernel
815 as a regular user if your Unix user is a member of the
816 <<tracing-group,tracing group>>.
818 . Create a <<tracing-session,tracing session>> which writes its traces
819 to dir:{/tmp/my-kernel-trace}:
824 # lttng create my-kernel-session --output=/tmp/my-kernel-trace
828 . List the available kernel tracepoints and system calls:
833 # lttng list --kernel
834 # lttng list --kernel --syscall
838 . Create <<event,event rules>> which match the desired instrumentation
839 point names, for example the `sched_switch` and `sched_process_fork`
840 tracepoints, and the man:open(2) and man:close(2) system calls:
845 # lttng enable-event --kernel sched_switch,sched_process_fork
846 # lttng enable-event --kernel --syscall open,close
850 You can also create an event rule which matches _all_ the Linux kernel
851 tracepoints (this will generate a lot of data when tracing):
856 # lttng enable-event --kernel --all
860 . <<basic-tracing-session-control,Start tracing>>:
869 . Do some operation on your system for a few seconds. For example,
870 load a website, or list the files of a directory.
871 . <<basic-tracing-session-control,Stop tracing>> and destroy the
882 You need to stop tracing to make LTTng flush the remaining trace data
883 and make the trace readable.
885 The man:lttng-destroy(1) command does not destroy the trace data; it
886 only destroys the state of the tracing session.
888 . For the sake of this example, make the recorded trace accessible to
894 # chown -R $(whoami) /tmp/my-kernel-trace
898 See <<viewing-and-analyzing-your-traces,View and analyze the
899 recorded events>> to view the recorded events.
902 [[tracing-your-own-user-application]]
903 === Trace a user application
905 This section steps you through a simple example to trace a
906 _Hello world_ program written in C.
908 To create the traceable user application:
910 . Create the tracepoint provider header file, which defines the
911 tracepoints and the events they can generate:
917 #undef TRACEPOINT_PROVIDER
918 #define TRACEPOINT_PROVIDER hello_world
920 #undef TRACEPOINT_INCLUDE
921 #define TRACEPOINT_INCLUDE "./hello-tp.h"
923 #if !defined(_HELLO_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
926 #include <lttng/tracepoint.h>
936 ctf_string(my_string_field, my_string_arg)
937 ctf_integer(int, my_integer_field, my_integer_arg)
941 #endif /* _HELLO_TP_H */
943 #include <lttng/tracepoint-event.h>
947 . Create the tracepoint provider package source file:
953 #define TRACEPOINT_CREATE_PROBES
954 #define TRACEPOINT_DEFINE
956 #include "hello-tp.h"
960 . Build the tracepoint provider package:
965 $ gcc -c -I. hello-tp.c
969 . Create the _Hello World_ application source file:
976 #include "hello-tp.h"
978 int main(int argc, char *argv[])
982 puts("Hello, World!\nPress Enter to continue...");
985 * The following getchar() call is only placed here for the purpose
986 * of this demonstration, to pause the application in order for
987 * you to have time to list its tracepoints. It is not
993 * A tracepoint() call.
995 * Arguments, as defined in hello-tp.h:
997 * 1. Tracepoint provider name (required)
998 * 2. Tracepoint name (required)
999 * 3. my_integer_arg (first user-defined argument)
1000 * 4. my_string_arg (second user-defined argument)
1002 * Notice the tracepoint provider and tracepoint names are
1003 * NOT strings: they are in fact parts of variables that the
1004 * macros in hello-tp.h create.
1006 tracepoint(hello_world, my_first_tracepoint, 23, "hi there!");
1008 for (x = 0; x < argc; ++x) {
1009 tracepoint(hello_world, my_first_tracepoint, x, argv[x]);
1012 puts("Quitting now!");
1013 tracepoint(hello_world, my_first_tracepoint, x * x, "x^2");
1020 . Build the application:
1029 . Link the application with the tracepoint provider package,
1030 `liblttng-ust`, and `libdl`:
1035 $ gcc -o hello hello.o hello-tp.o -llttng-ust -ldl
1039 Here's the whole build process:
1042 .User space tracing tutorial's build steps.
1043 image::ust-flow.png[]
1045 To trace the user application:
1047 . Run the application with a few arguments:
1052 $ ./hello world and beyond
1061 Press Enter to continue...
1065 . Start an LTTng <<lttng-sessiond,session daemon>>:
1070 $ lttng-sessiond --daemonize
1074 Note that a session daemon might already be running, for example as
1075 a service that the distribution's service manager started.
1077 . List the available user space tracepoints:
1082 $ lttng list --userspace
1086 You see the `hello_world:my_first_tracepoint` tracepoint listed
1087 under the `./hello` process.
1089 . Create a <<tracing-session,tracing session>>:
1094 $ lttng create my-user-space-session
1098 . Create an <<event,event rule>> which matches the
1099 `hello_world:my_first_tracepoint` event name:
1104 $ lttng enable-event --userspace hello_world:my_first_tracepoint
1108 . <<basic-tracing-session-control,Start tracing>>:
1117 . Go back to the running `hello` application and press Enter. The
1118 program executes all `tracepoint()` instrumentation points and exits.
1119 . <<basic-tracing-session-control,Stop tracing>> and destroy the
1130 You need to stop tracing to make LTTng flush the remaining trace data
1131 and make the trace readable.
1133 The man:lttng-destroy(1) command does not destroy the trace data; it
1134 only destroys the state of the tracing session.
1136 By default, LTTng saves the traces in
1137 +$LTTNG_HOME/lttng-traces/__name__-__date__-__time__+,
1138 where +__name__+ is the tracing session name. The
1139 env:LTTNG_HOME environment variable defaults to `$HOME` if not set.
1141 See <<viewing-and-analyzing-your-traces,View and analyze the
1142 recorded events>> to view the recorded events.
1145 [[viewing-and-analyzing-your-traces]]
1146 === View and analyze the recorded events
1148 Once you have completed the <<tracing-the-linux-kernel,Trace the Linux
1149 kernel>> and <<tracing-your-own-user-application,Trace a user
1150 application>> tutorials, you can inspect the recorded events.
1152 Many tools are available to read LTTng traces:
1154 * **cmd:babeltrace** is a command-line utility which converts trace
1155 formats; it supports the format that LTTng produces, CTF, as well as a
1156 basic text output which can be ++grep++ed. The cmd:babeltrace command
1157 is part of the http://diamon.org/babeltrace[Babeltrace] project.
1158 * Babeltrace also includes
1159 **https://www.python.org/[Python] bindings** so
1160 that you can easily open and read an LTTng trace with your own script,
1161 benefiting from the power of Python.
1162 * http://tracecompass.org/[**Trace Compass**]
1163 is a graphical user interface for viewing and analyzing any type of
1164 logs or traces, including LTTng's.
1165 * https://github.com/lttng/lttng-analyses[**LTTng analyses**] is a
1166 project which includes many high-level analyses of LTTng kernel
1167 traces, like scheduling statistics, interrupt frequency distribution,
1168 top CPU usage, and more.
1170 NOTE: This section assumes that the traces recorded during the previous
1171 tutorials were saved to their default location, in the
1172 dir:{$LTTNG_HOME/lttng-traces} directory. The env:LTTNG_HOME
1173 environment variable defaults to `$HOME` if not set.
1176 [[viewing-and-analyzing-your-traces-bt]]
1177 ==== Use the cmd:babeltrace command-line tool
1179 The simplest way to list all the recorded events of a trace is to pass
1180 its path to cmd:babeltrace with no options:
1184 $ babeltrace ~/lttng-traces/my-user-space-session*
1187 cmd:babeltrace finds all traces recursively within the given path and
1188 prints all their events, merging them in chronological order.
1190 You can pipe the output of cmd:babeltrace into a tool like man:grep(1) for
1195 $ babeltrace /tmp/my-kernel-trace | grep _switch
1198 You can pipe the output of cmd:babeltrace into a tool like man:wc(1) to
1199 count the recorded events:
1203 $ babeltrace /tmp/my-kernel-trace | grep _open | wc --lines
1207 [[viewing-and-analyzing-your-traces-bt-python]]
1208 ==== Use the Babeltrace Python bindings
1210 The <<viewing-and-analyzing-your-traces-bt,text output of cmd:babeltrace>>
1211 is useful to isolate events by simple matching using man:grep(1) and
1212 similar utilities. However, more elaborate filters, such as keeping only
1213 event records with a field value falling within a specific range, are
1214 not trivial to write using a shell. Moreover, reductions and even the
1215 most basic computations involving multiple event records are virtually
1216 impossible to implement.
1218 Fortunately, Babeltrace ships with Python 3 bindings which makes it easy
1219 to read the event records of an LTTng trace sequentially and compute the
1220 desired information.
1222 The following script accepts an LTTng Linux kernel trace path as its
1223 first argument and prints the short names of the top 5 running processes
1224 on CPU 0 during the whole trace:
1229 from collections import Counter
1235 if len(sys.argv) != 2:
1236 msg = 'Usage: python3 {} TRACEPATH'.format(sys.argv[0])
1237 print(msg, file=sys.stderr)
1240 # A trace collection contains one or more traces
1241 col = babeltrace.TraceCollection()
1243 # Add the trace provided by the user (LTTng traces always have
1245 if col.add_trace(sys.argv[1], 'ctf') is None:
1246 raise RuntimeError('Cannot add trace')
1248 # This counter dict contains execution times:
1250 # task command name -> total execution time (ns)
1251 exec_times = Counter()
1253 # This contains the last `sched_switch` timestamp
1257 for event in col.events:
1258 # Keep only `sched_switch` events
1259 if event.name != 'sched_switch':
1262 # Keep only events which happened on CPU 0
1263 if event['cpu_id'] != 0:
1267 cur_ts = event.timestamp
1273 # Previous task command (short) name
1274 prev_comm = event['prev_comm']
1276 # Initialize entry in our dict if not yet done
1277 if prev_comm not in exec_times:
1278 exec_times[prev_comm] = 0
1280 # Compute previous command execution time
1281 diff = cur_ts - last_ts
1283 # Update execution time of this command
1284 exec_times[prev_comm] += diff
1286 # Update last timestamp
1290 for name, ns in exec_times.most_common(5):
1292 print('{:20}{} s'.format(name, s))
1297 if __name__ == '__main__':
1298 sys.exit(0 if top5proc() else 1)
1305 $ python3 top5proc.py /tmp/my-kernel-trace/kernel
1311 swapper/0 48.607245889 s
1312 chromium 7.192738188 s
1313 pavucontrol 0.709894415 s
1314 Compositor 0.660867933 s
1315 Xorg.bin 0.616753786 s
1318 Note that `swapper/0` is the "idle" process of CPU 0 on Linux; since we
1319 weren't using the CPU that much when tracing, its first position in the
1324 == [[understanding-lttng]]Core concepts
1326 From a user's perspective, the LTTng system is built on a few concepts,
1327 or objects, on which the <<lttng-cli,cmd:lttng command-line tool>>
1328 operates by sending commands to the <<lttng-sessiond,session daemon>>.
1329 Understanding how those objects relate to eachother is key in mastering
1332 The core concepts are:
1334 * <<tracing-session,Tracing session>>
1335 * <<domain,Tracing domain>>
1336 * <<channel,Channel and ring buffer>>
1337 * <<"event","Instrumentation point, event rule, event, and event record">>
1343 A _tracing session_ is a stateful dialogue between you and
1344 a <<lttng-sessiond,session daemon>>. You can
1345 <<creating-destroying-tracing-sessions,create a new tracing
1346 session>> with the `lttng create` command.
1348 Anything that you do when you control LTTng tracers happens within a
1349 tracing session. In particular, a tracing session:
1352 * Has its own set of trace files.
1353 * Has its own state of activity (started or stopped).
1354 * Has its own <<tracing-session-mode,mode>> (local, network streaming,
1356 * Has its own <<channel,channels>> which have their own
1357 <<event,event rules>>.
1360 .A _tracing session_ contains <<channel,channels>> that are members of <<domain,tracing domains>> and contain <<event,event rules>>.
1361 image::concepts.png[]
1363 Those attributes and objects are completely isolated between different
1366 A tracing session is analogous to a cash machine session:
1367 the operations you do on the banking system through the cash machine do
1368 not alter the data of other users of the same system. In the case of
1369 the cash machine, a session lasts as long as your bank card is inside.
1370 In the case of LTTng, a tracing session lasts from the `lttng create`
1371 command to the `lttng destroy` command.
1374 .Each Unix user has its own set of tracing sessions.
1375 image::many-sessions.png[]
1378 [[tracing-session-mode]]
1379 ==== Tracing session mode
1381 LTTng can send the generated trace data to different locations. The
1382 _tracing session mode_ dictates where to send it. The following modes
1383 are available in LTTng{nbsp}{revision}:
1386 LTTng writes the traces to the file system of the machine being traced
1389 Network streaming mode::
1390 LTTng sends the traces over the network to a
1391 <<lttng-relayd,relay daemon>> running on a remote system.
1394 LTTng does not write the traces by default. Instead, you can request
1395 LTTng to <<taking-a-snapshot,take a snapshot>>, that is, a copy of the
1396 current tracing buffers, and to write it to the target's file system
1397 or to send it over the network to a <<lttng-relayd,relay daemon>>
1398 running on a remote system.
1401 This mode is similar to the network streaming mode, but a live
1402 trace viewer can connect to the distant relay daemon to
1403 <<lttng-live,view event records as LTTng generates them>> by
1410 A _tracing domain_ is a namespace for event sources. A tracing domain
1411 has its own properties and features.
1413 There are currently five available tracing domains:
1417 * `java.util.logging` (JUL)
1421 You must specify a tracing domain when using some commands to avoid
1422 ambiguity. For example, since all the domains support named tracepoints
1423 as event sources (instrumentation points that you manually insert in the
1424 source code), you need to specify a tracing domain when
1425 <<enabling-disabling-events,creating an event rule>> because all the
1426 tracing domains could have tracepoints with the same names.
1428 Some features are reserved to specific tracing domains. Dynamic function
1429 entry and return instrumentation points, for example, are currently only
1430 supported in the Linux kernel tracing domain, but support for other
1431 tracing domains could be added in the future.
1433 You can create <<channel,channels>> in the Linux kernel and user space
1434 tracing domains. The other tracing domains have a single default
1439 === Channel and ring buffer
1441 A _channel_ is an object which is responsible for a set of ring buffers.
1442 Each ring buffer is divided into multiple sub-buffers. When an LTTng
1443 tracer emits an event, it can record it to one or more
1444 sub-buffers. The attributes of a channel determine what to do when
1445 there's no space left for a new event record because all sub-buffers
1446 are full, where to send a full sub-buffer, and other behaviours.
1448 A channel is always associated to a <<domain,tracing domain>>. The
1449 `java.util.logging` (JUL), log4j, and Python tracing domains each have
1450 a default channel which you cannot configure.
1452 A channel also owns <<event,event rules>>. When an LTTng tracer emits
1453 an event, it records it to the sub-buffers of all
1454 the enabled channels with a satisfied event rule, as long as those
1455 channels are part of active <<tracing-session,tracing sessions>>.
1458 [[channel-buffering-schemes]]
1459 ==== Per-user vs. per-process buffering schemes
1461 A channel has at least one ring buffer _per CPU_. LTTng always
1462 records an event to the ring buffer associated to the CPU on which it
1465 Two _buffering schemes_ are available when you
1466 <<enabling-disabling-channels,create a channel>> in the
1467 user space <<domain,tracing domain>>:
1469 Per-user buffering::
1470 Allocate one set of ring buffers--one per CPU--shared by all the
1471 instrumented processes of each Unix user.
1475 .Per-user buffering scheme.
1476 image::per-user-buffering.png[]
1479 Per-process buffering::
1480 Allocate one set of ring buffers--one per CPU--for each
1481 instrumented process.
1485 .Per-process buffering scheme.
1486 image::per-process-buffering.png[]
1489 The per-process buffering scheme tends to consume more memory than the
1490 per-user option because systems generally have more instrumented
1491 processes than Unix users running instrumented processes. However, the
1492 per-process buffering scheme ensures that one process having a high
1493 event throughput won't fill all the shared sub-buffers of the same
1496 The Linux kernel tracing domain has only one available buffering scheme
1497 which is to allocate a single set of ring buffers for the whole system.
1498 This scheme is similar to the per-user option, but with a single, global
1499 user "running" the kernel.
1502 [[channel-overwrite-mode-vs-discard-mode]]
1503 ==== Overwrite vs. discard event loss modes
1505 When an event occurs, LTTng records it to a specific sub-buffer (yellow
1506 arc in the following animation) of a specific channel's ring buffer.
1507 When there's no space left in a sub-buffer, the tracer marks it as
1508 consumable (red) and another, empty sub-buffer starts receiving the
1509 following event records. A <<lttng-consumerd,consumer daemon>>
1510 eventually consumes the marked sub-buffer (returns to white).
1513 [role="docsvg-channel-subbuf-anim"]
1518 In an ideal world, sub-buffers are consumed faster than they are filled,
1519 as is the case in the previous animation. In the real world,
1520 however, all sub-buffers can be full at some point, leaving no space to
1521 record the following events.
1523 By default, LTTng-modules and LTTng-UST are _non-blocking_ tracers: when
1524 no empty sub-buffer is available, it is acceptable to lose event records
1525 when the alternative would be to cause substantial delays in the
1526 instrumented application's execution. LTTng privileges performance over
1527 integrity; it aims at perturbing the traced system as little as possible
1528 in order to make tracing of subtle race conditions and rare interrupt
1531 Starting from LTTng{nbsp}2.10, the LTTng user space tracer, LTTng-UST,
1532 supports a _blocking mode_. See the <<blocking-timeout-example,blocking
1533 timeout example>> to learn how to use the blocking mode.
1535 When it comes to losing event records because no empty sub-buffer is
1536 available, or because the <<opt-blocking-timeout,blocking timeout>> is
1537 reached, the channel's _event loss mode_ determines what to do. The
1538 available event loss modes are:
1541 Drop the newest event records until a the tracer releases a
1544 This is the only available mode when you specify a
1545 <<opt-blocking-timeout,blocking timeout>>.
1548 Clear the sub-buffer containing the oldest event records and start
1549 writing the newest event records there.
1551 This mode is sometimes called _flight recorder mode_ because it's
1553 https://en.wikipedia.org/wiki/Flight_recorder[flight recorder]:
1554 always keep a fixed amount of the latest data.
1556 Which mechanism you should choose depends on your context: prioritize
1557 the newest or the oldest event records in the ring buffer?
1559 Beware that, in overwrite mode, the tracer abandons a _whole sub-buffer_
1560 as soon as a there's no space left for a new event record, whereas in
1561 discard mode, the tracer only discards the event record that doesn't
1564 In discard mode, LTTng increments a count of lost event records when an
1565 event record is lost and saves this count to the trace. In overwrite
1566 mode, since LTTng 2.8, LTTng increments a count of lost sub-buffers when
1567 a sub-buffer is lost and saves this count to the trace. In this mode,
1568 the exact number of lost event records in those lost sub-buffers is not
1569 saved to the trace. Trace analyses can use the trace's saved discarded
1570 event record and sub-buffer counts to decide whether or not to perform
1571 the analyses even if trace data is known to be missing.
1573 There are a few ways to decrease your probability of losing event
1575 <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>> shows
1576 how you can fine-une the sub-buffer count and size of a channel to
1577 virtually stop losing event records, though at the cost of greater
1581 [[channel-subbuf-size-vs-subbuf-count]]
1582 ==== Sub-buffer count and size
1584 When you <<enabling-disabling-channels,create a channel>>, you can
1585 set its number of sub-buffers and their size.
1587 Note that there is noticeable CPU overhead introduced when
1588 switching sub-buffers (marking a full one as consumable and switching
1589 to an empty one for the following events to be recorded). Knowing this,
1590 the following list presents a few practical situations along with how
1591 to configure the sub-buffer count and size for them:
1593 * **High event throughput**: In general, prefer bigger sub-buffers to
1594 lower the risk of losing event records.
1596 Having bigger sub-buffers also ensures a lower
1597 <<channel-switch-timer,sub-buffer switching frequency>>.
1599 The number of sub-buffers is only meaningful if you create the channel
1600 in overwrite mode: in this case, if a sub-buffer overwrite happens, the
1601 other sub-buffers are left unaltered.
1603 * **Low event throughput**: In general, prefer smaller sub-buffers
1604 since the risk of losing event records is low.
1606 Because events occur less frequently, the sub-buffer switching frequency
1607 should remain low and thus the tracer's overhead should not be a
1610 * **Low memory system**: If your target system has a low memory
1611 limit, prefer fewer first, then smaller sub-buffers.
1613 Even if the system is limited in memory, you want to keep the
1614 sub-buffers as big as possible to avoid a high sub-buffer switching
1617 Note that LTTng uses http://diamon.org/ctf/[CTF] as its trace format,
1618 which means event data is very compact. For example, the average
1619 LTTng kernel event record weights about 32{nbsp}bytes. Thus, a
1620 sub-buffer size of 1{nbsp}MiB is considered big.
1622 The previous situations highlight the major trade-off between a few big
1623 sub-buffers and more, smaller sub-buffers: sub-buffer switching
1624 frequency vs. how much data is lost in overwrite mode. Assuming a
1625 constant event throughput and using the overwrite mode, the two
1626 following configurations have the same ring buffer total size:
1629 [role="docsvg-channel-subbuf-size-vs-count-anim"]
1634 * **2 sub-buffers of 4{nbsp}MiB each**: Expect a very low sub-buffer
1635 switching frequency, but if a sub-buffer overwrite happens, half of
1636 the event records so far (4{nbsp}MiB) are definitely lost.
1637 * **8 sub-buffers of 1{nbsp}MiB each**: Expect 4{nbsp}times the tracer's
1638 overhead as the previous configuration, but if a sub-buffer
1639 overwrite happens, only the eighth of event records so far are
1642 In discard mode, the sub-buffers count parameter is pointless: use two
1643 sub-buffers and set their size according to the requirements of your
1647 [[channel-switch-timer]]
1648 ==== Switch timer period
1650 The _switch timer period_ is an important configurable attribute of
1651 a channel to ensure periodic sub-buffer flushing.
1653 When the _switch timer_ expires, a sub-buffer switch happens. You can
1654 set the switch timer period attribute when you
1655 <<enabling-disabling-channels,create a channel>> to ensure that event
1656 data is consumed and committed to trace files or to a distant relay
1657 daemon periodically in case of a low event throughput.
1660 [role="docsvg-channel-switch-timer"]
1665 This attribute is also convenient when you use big sub-buffers to cope
1666 with a sporadic high event throughput, even if the throughput is
1670 [[channel-read-timer]]
1671 ==== Read timer period
1673 By default, the LTTng tracers use a notification mechanism to signal a
1674 full sub-buffer so that a consumer daemon can consume it. When such
1675 notifications must be avoided, for example in real-time applications,
1676 you can use the channel's _read timer_ instead. When the read timer
1677 fires, the <<lttng-consumerd,consumer daemon>> checks for full,
1678 consumable sub-buffers.
1681 [[tracefile-rotation]]
1682 ==== Trace file count and size
1684 By default, trace files can grow as large as needed. You can set the
1685 maximum size of each trace file that a channel writes when you
1686 <<enabling-disabling-channels,create a channel>>. When the size of
1687 a trace file reaches the channel's fixed maximum size, LTTng creates
1688 another file to contain the next event records. LTTng appends a file
1689 count to each trace file name in this case.
1691 If you set the trace file size attribute when you create a channel, the
1692 maximum number of trace files that LTTng creates is _unlimited_ by
1693 default. To limit them, you can also set a maximum number of trace
1694 files. When the number of trace files reaches the channel's fixed
1695 maximum count, the oldest trace file is overwritten. This mechanism is
1696 called _trace file rotation_.
1700 === Instrumentation point, event rule, event, and event record
1702 An _event rule_ is a set of conditions which must be **all** satisfied
1703 for LTTng to record an occuring event.
1705 You set the conditions when you <<enabling-disabling-events,create
1708 You always attach an event rule to <<channel,channel>> when you create
1711 When an event passes the conditions of an event rule, LTTng records it
1712 in one of the attached channel's sub-buffers.
1714 The available conditions, as of LTTng{nbsp}{revision}, are:
1716 * The event rule _is enabled_.
1717 * The instrumentation point's type _is{nbsp}T_.
1718 * The instrumentation point's name (sometimes called _event name_)
1719 _matches{nbsp}N_, but _is not{nbsp}E_.
1720 * The instrumentation point's log level _is as severe as{nbsp}L_, or
1721 _is exactly{nbsp}L_.
1722 * The fields of the event's payload _satisfy_ a filter
1723 expression{nbsp}__F__.
1725 As you can see, all the conditions but the dynamic filter are related to
1726 the event rule's status or to the instrumentation point, not to the
1727 occurring events. This is why, without a filter, checking if an event
1728 passes an event rule is not a dynamic task: when you create or modify an
1729 event rule, all the tracers of its tracing domain enable or disable the
1730 instrumentation points themselves once. This is possible because the
1731 attributes of an instrumentation point (type, name, and log level) are
1732 defined statically. In other words, without a dynamic filter, the tracer
1733 _does not evaluate_ the arguments of an instrumentation point unless it
1734 matches an enabled event rule.
1736 Note that, for LTTng to record an event, the <<channel,channel>> to
1737 which a matching event rule is attached must also be enabled, and the
1738 tracing session owning this channel must be active.
1741 .Logical path from an instrumentation point to an event record.
1742 image::event-rule.png[]
1744 .Event, event record, or event rule?
1746 With so many similar terms, it's easy to get confused.
1748 An **event** is the consequence of the execution of an _instrumentation
1749 point_, like a tracepoint that you manually place in some source code,
1750 or a Linux kernel KProbe. An event is said to _occur_ at a specific
1751 time. Different actions can be taken upon the occurrence of an event,
1752 like record the event's payload to a buffer.
1754 An **event record** is the representation of an event in a sub-buffer. A
1755 tracer is responsible for capturing the payload of an event, current
1756 context variables, the event's ID, and the event's timestamp. LTTng
1757 can append this sub-buffer to a trace file.
1759 An **event rule** is a set of conditions which must all be satisfied for
1760 LTTng to record an occuring event. Events still occur without
1761 satisfying event rules, but LTTng does not record them.
1766 == Components of noch:{LTTng}
1768 The second _T_ in _LTTng_ stands for _toolkit_: it would be wrong
1769 to call LTTng a simple _tool_ since it is composed of multiple
1770 interacting components. This section describes those components,
1771 explains their respective roles, and shows how they connect together to
1772 form the LTTng ecosystem.
1774 The following diagram shows how the most important components of LTTng
1775 interact with user applications, the Linux kernel, and you:
1778 .Control and trace data paths between LTTng components.
1779 image::plumbing.png[]
1781 The LTTng project incorporates:
1783 * **LTTng-tools**: Libraries and command-line interface to
1784 control tracing sessions.
1785 ** <<lttng-sessiond,Session daemon>> (man:lttng-sessiond(8)).
1786 ** <<lttng-consumerd,Consumer daemon>> (cmd:lttng-consumerd).
1787 ** <<lttng-relayd,Relay daemon>> (man:lttng-relayd(8)).
1788 ** <<liblttng-ctl-lttng,Tracing control library>> (`liblttng-ctl`).
1789 ** <<lttng-cli,Tracing control command-line tool>> (man:lttng(1)).
1790 * **LTTng-UST**: Libraries and Java/Python packages to trace user
1792 ** <<lttng-ust,User space tracing library>> (`liblttng-ust`) and its
1793 headers to instrument and trace any native user application.
1794 ** <<prebuilt-ust-helpers,Preloadable user space tracing helpers>>:
1795 *** `liblttng-ust-libc-wrapper`
1796 *** `liblttng-ust-pthread-wrapper`
1797 *** `liblttng-ust-cyg-profile`
1798 *** `liblttng-ust-cyg-profile-fast`
1799 *** `liblttng-ust-dl`
1800 ** User space tracepoint provider source files generator command-line
1801 tool (man:lttng-gen-tp(1)).
1802 ** <<lttng-ust-agents,LTTng-UST Java agent>> to instrument and trace
1803 Java applications using `java.util.logging` or
1804 Apache log4j 1.2 logging.
1805 ** <<lttng-ust-agents,LTTng-UST Python agent>> to instrument
1806 Python applications using the standard `logging` package.
1807 * **LTTng-modules**: <<lttng-modules,Linux kernel modules>> to trace
1809 ** LTTng kernel tracer module.
1810 ** Tracing ring buffer kernel modules.
1811 ** Probe kernel modules.
1812 ** LTTng logger kernel module.
1816 === Tracing control command-line interface
1819 .The tracing control command-line interface.
1820 image::plumbing-lttng-cli.png[]
1822 The _man:lttng(1) command-line tool_ is the standard user interface to
1823 control LTTng <<tracing-session,tracing sessions>>. The cmd:lttng tool
1824 is part of LTTng-tools.
1826 The cmd:lttng tool is linked with
1827 <<liblttng-ctl-lttng,`liblttng-ctl`>> to communicate with
1828 one or more <<lttng-sessiond,session daemons>> behind the scenes.
1830 The cmd:lttng tool has a Git-like interface:
1834 $ lttng <GENERAL OPTIONS> <COMMAND> <COMMAND OPTIONS>
1837 The <<controlling-tracing,Tracing control>> section explores the
1838 available features of LTTng using the cmd:lttng tool.
1841 [[liblttng-ctl-lttng]]
1842 === Tracing control library
1845 .The tracing control library.
1846 image::plumbing-liblttng-ctl.png[]
1848 The _LTTng control library_, `liblttng-ctl`, is used to communicate
1849 with a <<lttng-sessiond,session daemon>> using a C API that hides the
1850 underlying protocol's details. `liblttng-ctl` is part of LTTng-tools.
1852 The <<lttng-cli,cmd:lttng command-line tool>>
1853 is linked with `liblttng-ctl`.
1855 You can use `liblttng-ctl` in C or $$C++$$ source code by including its
1860 #include <lttng/lttng.h>
1863 Some objects are referenced by name (C string), such as tracing
1864 sessions, but most of them require to create a handle first using
1865 `lttng_create_handle()`.
1867 The best available developer documentation for `liblttng-ctl` is, as of
1868 LTTng{nbsp}{revision}, its installed header files. Every function and
1869 structure is thoroughly documented.
1873 === User space tracing library
1876 .The user space tracing library.
1877 image::plumbing-liblttng-ust.png[]
1879 The _user space tracing library_, `liblttng-ust` (see man:lttng-ust(3)),
1880 is the LTTng user space tracer. It receives commands from a
1881 <<lttng-sessiond,session daemon>>, for example to
1882 enable and disable specific instrumentation points, and writes event
1883 records to ring buffers shared with a
1884 <<lttng-consumerd,consumer daemon>>.
1885 `liblttng-ust` is part of LTTng-UST.
1887 Public C header files are installed beside `liblttng-ust` to
1888 instrument any <<c-application,C or $$C++$$ application>>.
1890 <<lttng-ust-agents,LTTng-UST agents>>, which are regular Java and Python
1891 packages, use their own library providing tracepoints which is
1892 linked with `liblttng-ust`.
1894 An application or library does not have to initialize `liblttng-ust`
1895 manually: its constructor does the necessary tasks to properly register
1896 to a session daemon. The initialization phase also enables the
1897 instrumentation points matching the <<event,event rules>> that you
1901 [[lttng-ust-agents]]
1902 === User space tracing agents
1905 .The user space tracing agents.
1906 image::plumbing-lttng-ust-agents.png[]
1908 The _LTTng-UST Java and Python agents_ are regular Java and Python
1909 packages which add LTTng tracing capabilities to the
1910 native logging frameworks. The LTTng-UST agents are part of LTTng-UST.
1912 In the case of Java, the
1913 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[`java.util.logging`
1914 core logging facilities] and
1915 https://logging.apache.org/log4j/1.2/[Apache log4j 1.2] are supported.
1916 Note that Apache Log4{nbsp}2 is not supported.
1918 In the case of Python, the standard
1919 https://docs.python.org/3/library/logging.html[`logging`] package
1920 is supported. Both Python 2 and Python 3 modules can import the
1921 LTTng-UST Python agent package.
1923 The applications using the LTTng-UST agents are in the
1924 `java.util.logging` (JUL),
1925 log4j, and Python <<domain,tracing domains>>.
1927 Both agents use the same mechanism to trace the log statements. When an
1928 agent is initialized, it creates a log handler that attaches to the root
1929 logger. The agent also registers to a <<lttng-sessiond,session daemon>>.
1930 When the application executes a log statement, it is passed to the
1931 agent's log handler by the root logger. The agent's log handler calls a
1932 native function in a tracepoint provider package shared library linked
1933 with <<lttng-ust,`liblttng-ust`>>, passing the formatted log message and
1934 other fields, like its logger name and its log level. This native
1935 function contains a user space instrumentation point, hence tracing the
1938 The log level condition of an
1939 <<event,event rule>> is considered when tracing
1940 a Java or a Python application, and it's compatible with the standard
1941 JUL, log4j, and Python log levels.
1945 === LTTng kernel modules
1948 .The LTTng kernel modules.
1949 image::plumbing-lttng-modules.png[]
1951 The _LTTng kernel modules_ are a set of Linux kernel modules
1952 which implement the kernel tracer of the LTTng project. The LTTng
1953 kernel modules are part of LTTng-modules.
1955 The LTTng kernel modules include:
1957 * A set of _probe_ modules.
1959 Each module attaches to a specific subsystem
1960 of the Linux kernel using its tracepoint instrument points. There are
1961 also modules to attach to the entry and return points of the Linux
1962 system call functions.
1964 * _Ring buffer_ modules.
1966 A ring buffer implementation is provided as kernel modules. The LTTng
1967 kernel tracer writes to the ring buffer; a
1968 <<lttng-consumerd,consumer daemon>> reads from the ring buffer.
1970 * The _LTTng kernel tracer_ module.
1971 * The _LTTng logger_ module.
1973 The LTTng logger module implements the special path:{/proc/lttng-logger}
1974 file so that any executable can generate LTTng events by opening and
1975 writing to this file.
1977 See <<proc-lttng-logger-abi,LTTng logger>>.
1979 Generally, you do not have to load the LTTng kernel modules manually
1980 (using man:modprobe(8), for example): a root <<lttng-sessiond,session
1981 daemon>> loads the necessary modules when starting. If you have extra
1982 probe modules, you can specify to load them to the session daemon on
1985 The LTTng kernel modules are installed in
1986 +/usr/lib/modules/__release__/extra+ by default, where +__release__+ is
1987 the kernel release (see `uname --kernel-release`).
1994 .The session daemon.
1995 image::plumbing-sessiond.png[]
1997 The _session daemon_, man:lttng-sessiond(8), is a daemon responsible for
1998 managing tracing sessions and for controlling the various components of
1999 LTTng. The session daemon is part of LTTng-tools.
2001 The session daemon sends control requests to and receives control
2004 * The <<lttng-ust,user space tracing library>>.
2006 Any instance of the user space tracing library first registers to
2007 a session daemon. Then, the session daemon can send requests to
2008 this instance, such as:
2011 ** Get the list of tracepoints.
2012 ** Share an <<event,event rule>> so that the user space tracing library
2013 can enable or disable tracepoints. Amongst the possible conditions
2014 of an event rule is a filter expression which `liblttng-ust` evalutes
2015 when an event occurs.
2016 ** Share <<channel,channel>> attributes and ring buffer locations.
2019 The session daemon and the user space tracing library use a Unix
2020 domain socket for their communication.
2022 * The <<lttng-ust-agents,user space tracing agents>>.
2024 Any instance of a user space tracing agent first registers to
2025 a session daemon. Then, the session daemon can send requests to
2026 this instance, such as:
2029 ** Get the list of loggers.
2030 ** Enable or disable a specific logger.
2033 The session daemon and the user space tracing agent use a TCP connection
2034 for their communication.
2036 * The <<lttng-modules,LTTng kernel tracer>>.
2037 * The <<lttng-consumerd,consumer daemon>>.
2039 The session daemon sends requests to the consumer daemon to instruct
2040 it where to send the trace data streams, amongst other information.
2042 * The <<lttng-relayd,relay daemon>>.
2044 The session daemon receives commands from the
2045 <<liblttng-ctl-lttng,tracing control library>>.
2047 The root session daemon loads the appropriate
2048 <<lttng-modules,LTTng kernel modules>> on startup. It also spawns
2049 a <<lttng-consumerd,consumer daemon>> as soon as you create
2050 an <<event,event rule>>.
2052 The session daemon does not send and receive trace data: this is the
2053 role of the <<lttng-consumerd,consumer daemon>> and
2054 <<lttng-relayd,relay daemon>>. It does, however, generate the
2055 http://diamon.org/ctf/[CTF] metadata stream.
2057 Each Unix user can have its own session daemon instance. The
2058 tracing sessions managed by different session daemons are completely
2061 The root user's session daemon is the only one which is
2062 allowed to control the LTTng kernel tracer, and its spawned consumer
2063 daemon is the only one which is allowed to consume trace data from the
2064 LTTng kernel tracer. Note, however, that any Unix user which is a member
2065 of the <<tracing-group,tracing group>> is allowed
2066 to create <<channel,channels>> in the
2067 Linux kernel <<domain,tracing domain>>, and thus to trace the Linux
2070 The <<lttng-cli,cmd:lttng command-line tool>> automatically starts a
2071 session daemon when using its `create` command if none is currently
2072 running. You can also start the session daemon manually.
2079 .The consumer daemon.
2080 image::plumbing-consumerd.png[]
2082 The _consumer daemon_, cmd:lttng-consumerd, is a daemon which shares
2083 ring buffers with user applications or with the LTTng kernel modules to
2084 collect trace data and send it to some location (on disk or to a
2085 <<lttng-relayd,relay daemon>> over the network). The consumer daemon
2086 is part of LTTng-tools.
2088 You do not start a consumer daemon manually: a consumer daemon is always
2089 spawned by a <<lttng-sessiond,session daemon>> as soon as you create an
2090 <<event,event rule>>, that is, before you start tracing. When you kill
2091 its owner session daemon, the consumer daemon also exits because it is
2092 the session daemon's child process. Command-line options of
2093 man:lttng-sessiond(8) target the consumer daemon process.
2095 There are up to two running consumer daemons per Unix user, whereas only
2096 one session daemon can run per user. This is because each process can be
2097 either 32-bit or 64-bit: if the target system runs a mixture of 32-bit
2098 and 64-bit processes, it is more efficient to have separate
2099 corresponding 32-bit and 64-bit consumer daemons. The root user is an
2100 exception: it can have up to _three_ running consumer daemons: 32-bit
2101 and 64-bit instances for its user applications, and one more
2102 reserved for collecting kernel trace data.
2110 image::plumbing-relayd.png[]
2112 The _relay daemon_, man:lttng-relayd(8), is a daemon acting as a bridge
2113 between remote session and consumer daemons, local trace files, and a
2114 remote live trace viewer. The relay daemon is part of LTTng-tools.
2116 The main purpose of the relay daemon is to implement a receiver of
2117 <<sending-trace-data-over-the-network,trace data over the network>>.
2118 This is useful when the target system does not have much file system
2119 space to record trace files locally.
2121 The relay daemon is also a server to which a
2122 <<lttng-live,live trace viewer>> can
2123 connect. The live trace viewer sends requests to the relay daemon to
2124 receive trace data as the target system emits events. The
2125 communication protocol is named _LTTng live_; it is used over TCP
2128 Note that you can start the relay daemon on the target system directly.
2129 This is the setup of choice when the use case is to view events as
2130 the target system emits them without the need of a remote system.
2134 == [[using-lttng]]Instrumentation
2136 There are many examples of tracing and monitoring in our everyday life:
2138 * You have access to real-time and historical weather reports and
2139 forecasts thanks to weather stations installed around the country.
2140 * You know your heart is safe thanks to an electrocardiogram.
2141 * You make sure not to drive your car too fast and to have enough fuel
2142 to reach your destination thanks to gauges visible on your dashboard.
2144 All the previous examples have something in common: they rely on
2145 **instruments**. Without the electrodes attached to the surface of your
2146 body's skin, cardiac monitoring is futile.
2148 LTTng, as a tracer, is no different from those real life examples. If
2149 you're about to trace a software system or, in other words, record its
2150 history of execution, you better have **instrumentation points** in the
2151 subject you're tracing, that is, the actual software.
2153 Various ways were developed to instrument a piece of software for LTTng
2154 tracing. The most straightforward one is to manually place
2155 instrumentation points, called _tracepoints_, in the software's source
2156 code. It is also possible to add instrumentation points dynamically in
2157 the Linux kernel <<domain,tracing domain>>.
2159 If you're only interested in tracing the Linux kernel, your
2160 instrumentation needs are probably already covered by LTTng's built-in
2161 <<lttng-modules,Linux kernel tracepoints>>. You may also wish to trace a
2162 user application which is already instrumented for LTTng tracing.
2163 In such cases, you can skip this whole section and read the topics of
2164 the <<controlling-tracing,Tracing control>> section.
2166 Many methods are available to instrument a piece of software for LTTng
2169 * <<c-application,User space instrumentation for C and $$C++$$
2171 * <<prebuilt-ust-helpers,Prebuilt user space tracing helpers>>.
2172 * <<java-application,User space Java agent>>.
2173 * <<python-application,User space Python agent>>.
2174 * <<proc-lttng-logger-abi,LTTng logger>>.
2175 * <<instrumenting-linux-kernel,LTTng kernel tracepoints>>.
2179 === [[cxx-application]]User space instrumentation for C and $$C++$$ applications
2181 The procedure to instrument a C or $$C++$$ user application with
2182 the <<lttng-ust,LTTng user space tracing library>>, `liblttng-ust`, is:
2184 . <<tracepoint-provider,Create the source files of a tracepoint provider
2186 . <<probing-the-application-source-code,Add tracepoints to
2187 the application's source code>>.
2188 . <<building-tracepoint-providers-and-user-application,Build and link
2189 a tracepoint provider package and the user application>>.
2191 If you need quick, man:printf(3)-like instrumentation, you can skip
2192 those steps and use <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>>
2195 IMPORTANT: You need to <<installing-lttng,install>> LTTng-UST to
2196 instrument a user application with `liblttng-ust`.
2199 [[tracepoint-provider]]
2200 ==== Create the source files of a tracepoint provider package
2202 A _tracepoint provider_ is a set of compiled functions which provide
2203 **tracepoints** to an application, the type of instrumentation point
2204 supported by LTTng-UST. Those functions can emit events with
2205 user-defined fields and serialize those events as event records to one
2206 or more LTTng-UST <<channel,channel>> sub-buffers. The `tracepoint()`
2207 macro, which you <<probing-the-application-source-code,insert in a user
2208 application's source code>>, calls those functions.
2210 A _tracepoint provider package_ is an object file (`.o`) or a shared
2211 library (`.so`) which contains one or more tracepoint providers.
2212 Its source files are:
2214 * One or more <<tpp-header,tracepoint provider header>> (`.h`).
2215 * A <<tpp-source,tracepoint provider package source>> (`.c`).
2217 A tracepoint provider package is dynamically linked with `liblttng-ust`,
2218 the LTTng user space tracer, at run time.
2221 .User application linked with `liblttng-ust` and containing a tracepoint provider.
2222 image::ust-app.png[]
2224 NOTE: If you need quick, man:printf(3)-like instrumentation, you can
2225 skip creating and using a tracepoint provider and use
2226 <<tracef,`tracef()`>> or <<tracelog,`tracelog()`>> instead.
2230 ===== Create a tracepoint provider header file template
2232 A _tracepoint provider header file_ contains the tracepoint
2233 definitions of a tracepoint provider.
2235 To create a tracepoint provider header file:
2237 . Start from this template:
2241 .Tracepoint provider header file template (`.h` file extension).
2243 #undef TRACEPOINT_PROVIDER
2244 #define TRACEPOINT_PROVIDER provider_name
2246 #undef TRACEPOINT_INCLUDE
2247 #define TRACEPOINT_INCLUDE "./tp.h"
2249 #if !defined(_TP_H) || defined(TRACEPOINT_HEADER_MULTI_READ)
2252 #include <lttng/tracepoint.h>
2255 * Use TRACEPOINT_EVENT(), TRACEPOINT_EVENT_CLASS(),
2256 * TRACEPOINT_EVENT_INSTANCE(), and TRACEPOINT_LOGLEVEL() here.
2261 #include <lttng/tracepoint-event.h>
2267 * `provider_name` with the name of your tracepoint provider.
2268 * `"tp.h"` with the name of your tracepoint provider header file.
2270 . Below the `#include <lttng/tracepoint.h>` line, put your
2271 <<defining-tracepoints,tracepoint definitions>>.
2273 Your tracepoint provider name must be unique amongst all the possible
2274 tracepoint provider names used on the same target system. We
2275 suggest to include the name of your project or company in the name,
2276 for example, `org_lttng_my_project_tpp`.
2278 TIP: [[lttng-gen-tp]]You can use the man:lttng-gen-tp(1) tool to create
2279 this boilerplate for you. When using cmd:lttng-gen-tp, all you need to
2280 write are the <<defining-tracepoints,tracepoint definitions>>.
2283 [[defining-tracepoints]]
2284 ===== Create a tracepoint definition
2286 A _tracepoint definition_ defines, for a given tracepoint:
2288 * Its **input arguments**. They are the macro parameters that the
2289 `tracepoint()` macro accepts for this particular tracepoint
2290 in the user application's source code.
2291 * Its **output event fields**. They are the sources of event fields
2292 that form the payload of any event that the execution of the
2293 `tracepoint()` macro emits for this particular tracepoint.
2295 You can create a tracepoint definition by using the
2296 `TRACEPOINT_EVENT()` macro below the `#include <lttng/tracepoint.h>`
2298 <<tpp-header,tracepoint provider header file template>>.
2300 The syntax of the `TRACEPOINT_EVENT()` macro is:
2303 .`TRACEPOINT_EVENT()` macro syntax.
2306 /* Tracepoint provider name */
2309 /* Tracepoint name */
2312 /* Input arguments */
2317 /* Output event fields */
2326 * `provider_name` with your tracepoint provider name.
2327 * `tracepoint_name` with your tracepoint name.
2328 * `arguments` with the <<tpp-def-input-args,input arguments>>.
2329 * `fields` with the <<tpp-def-output-fields,output event field>>
2332 This tracepoint emits events named `provider_name:tracepoint_name`.
2335 .Event name's length limitation
2337 The concatenation of the tracepoint provider name and the
2338 tracepoint name must not exceed **254 characters**. If it does, the
2339 instrumented application compiles and runs, but LTTng throws multiple
2340 warnings and you could experience serious issues.
2343 [[tpp-def-input-args]]The syntax of the `TP_ARGS()` macro is:
2346 .`TP_ARGS()` macro syntax.
2355 * `type` with the C type of the argument.
2356 * `arg_name` with the argument name.
2358 You can repeat `type` and `arg_name` up to 10 times to have
2359 more than one argument.
2361 .`TP_ARGS()` usage with three arguments.
2373 The `TP_ARGS()` and `TP_ARGS(void)` forms are valid to create a
2374 tracepoint definition with no input arguments.
2376 [[tpp-def-output-fields]]The `TP_FIELDS()` macro contains a list of
2377 `ctf_*()` macros. Each `ctf_*()` macro defines one event field. See
2378 man:lttng-ust(3) for a complete description of the available `ctf_*()`
2379 macros. A `ctf_*()` macro specifies the type, size, and byte order of
2382 Each `ctf_*()` macro takes an _argument expression_ parameter. This is a
2383 C expression that the tracer evalutes at the `tracepoint()` macro site
2384 in the application's source code. This expression provides a field's
2385 source of data. The argument expression can include input argument names
2386 listed in the `TP_ARGS()` macro.
2388 Each `ctf_*()` macro also takes a _field name_ parameter. Field names
2389 must be unique within a given tracepoint definition.
2391 Here's a complete tracepoint definition example:
2393 .Tracepoint definition.
2395 The following tracepoint definition defines a tracepoint which takes
2396 three input arguments and has four output event fields.
2400 #include "my-custom-structure.h"
2406 const struct my_custom_structure*, my_custom_structure,
2411 ctf_string(query_field, query)
2412 ctf_float(double, ratio_field, ratio)
2413 ctf_integer(int, recv_size, my_custom_structure->recv_size)
2414 ctf_integer(int, send_size, my_custom_structure->send_size)
2419 You can refer to this tracepoint definition with the `tracepoint()`
2420 macro in your application's source code like this:
2424 tracepoint(my_provider, my_tracepoint,
2425 my_structure, some_ratio, the_query);
2429 NOTE: The LTTng tracer only evaluates tracepoint arguments at run time
2430 if they satisfy an enabled <<event,event rule>>.
2433 [[using-tracepoint-classes]]
2434 ===== Use a tracepoint class
2436 A _tracepoint class_ is a class of tracepoints which share the same
2437 output event field definitions. A _tracepoint instance_ is one
2438 instance of such a defined tracepoint class, with its own tracepoint
2441 The <<defining-tracepoints,`TRACEPOINT_EVENT()` macro>> is actually a
2442 shorthand which defines both a tracepoint class and a tracepoint
2443 instance at the same time.
2445 When you build a tracepoint provider package, the C or $$C++$$ compiler
2446 creates one serialization function for each **tracepoint class**. A
2447 serialization function is responsible for serializing the event fields
2448 of a tracepoint to a sub-buffer when tracing.
2450 For various performance reasons, when your situation requires multiple
2451 tracepoint definitions with different names, but with the same event
2452 fields, we recommend that you manually create a tracepoint class
2453 and instantiate as many tracepoint instances as needed. One positive
2454 effect of such a design, amongst other advantages, is that all
2455 tracepoint instances of the same tracepoint class reuse the same
2456 serialization function, thus reducing
2457 https://en.wikipedia.org/wiki/Cache_pollution[cache pollution].
2459 .Use a tracepoint class and tracepoint instances.
2461 Consider the following three tracepoint definitions:
2473 ctf_integer(int, userid, userid)
2474 ctf_integer(size_t, len, len)
2486 ctf_integer(int, userid, userid)
2487 ctf_integer(size_t, len, len)
2499 ctf_integer(int, userid, userid)
2500 ctf_integer(size_t, len, len)
2505 In this case, we create three tracepoint classes, with one implicit
2506 tracepoint instance for each of them: `get_account`, `get_settings`, and
2507 `get_transaction`. However, they all share the same event field names
2508 and types. Hence three identical, yet independent serialization
2509 functions are created when you build the tracepoint provider package.
2511 A better design choice is to define a single tracepoint class and three
2512 tracepoint instances:
2516 /* The tracepoint class */
2517 TRACEPOINT_EVENT_CLASS(
2518 /* Tracepoint provider name */
2521 /* Tracepoint class name */
2524 /* Input arguments */
2530 /* Output event fields */
2532 ctf_integer(int, userid, userid)
2533 ctf_integer(size_t, len, len)
2537 /* The tracepoint instances */
2538 TRACEPOINT_EVENT_INSTANCE(
2539 /* Tracepoint provider name */
2542 /* Tracepoint class name */
2545 /* Tracepoint name */
2548 /* Input arguments */
2554 TRACEPOINT_EVENT_INSTANCE(
2563 TRACEPOINT_EVENT_INSTANCE(
2576 [[assigning-log-levels]]
2577 ===== Assign a log level to a tracepoint definition
2579 You can assign an optional _log level_ to a
2580 <<defining-tracepoints,tracepoint definition>>.
2582 Assigning different levels of severity to tracepoint definitions can
2583 be useful: when you <<enabling-disabling-events,create an event rule>>,
2584 you can target tracepoints having a log level as severe as a specific
2587 The concept of LTTng-UST log levels is similar to the levels found
2588 in typical logging frameworks:
2590 * In a logging framework, the log level is given by the function
2591 or method name you use at the log statement site: `debug()`,
2592 `info()`, `warn()`, `error()`, and so on.
2593 * In LTTng-UST, you statically assign the log level to a tracepoint
2594 definition; any `tracepoint()` macro invocation which refers to
2595 this definition has this log level.
2597 You can assign a log level to a tracepoint definition with the
2598 `TRACEPOINT_LOGLEVEL()` macro. You must use this macro _after_ the
2599 <<defining-tracepoints,`TRACEPOINT_EVENT()`>> or
2600 <<using-tracepoint-classes,`TRACEPOINT_INSTANCE()`>> macro for a given
2603 The syntax of the `TRACEPOINT_LOGLEVEL()` macro is:
2606 .`TRACEPOINT_LOGLEVEL()` macro syntax.
2608 TRACEPOINT_LOGLEVEL(provider_name, tracepoint_name, log_level)
2613 * `provider_name` with the tracepoint provider name.
2614 * `tracepoint_name` with the tracepoint name.
2615 * `log_level` with the log level to assign to the tracepoint
2616 definition named `tracepoint_name` in the `provider_name`
2617 tracepoint provider.
2619 See man:lttng-ust(3) for a list of available log level names.
2621 .Assign the `TRACE_DEBUG_UNIT` log level to a tracepoint definition.
2625 /* Tracepoint definition */
2634 ctf_integer(int, userid, userid)
2635 ctf_integer(size_t, len, len)
2639 /* Log level assignment */
2640 TRACEPOINT_LOGLEVEL(my_app, get_transaction, TRACE_DEBUG_UNIT)
2646 ===== Create a tracepoint provider package source file
2648 A _tracepoint provider package source file_ is a C source file which
2649 includes a <<tpp-header,tracepoint provider header file>> to expand its
2650 macros into event serialization and other functions.
2652 You can always use the following tracepoint provider package source
2656 .Tracepoint provider package source file template.
2658 #define TRACEPOINT_CREATE_PROBES
2663 Replace `tp.h` with the name of your <<tpp-header,tracepoint provider
2664 header file>> name. You may also include more than one tracepoint
2665 provider header file here to create a tracepoint provider package
2666 holding more than one tracepoint providers.
2669 [[probing-the-application-source-code]]
2670 ==== Add tracepoints to an application's source code
2672 Once you <<tpp-header,create a tracepoint provider header file>>, you
2673 can use the `tracepoint()` macro in your application's
2674 source code to insert the tracepoints that this header
2675 <<defining-tracepoints,defines>>.
2677 The `tracepoint()` macro takes at least two parameters: the tracepoint
2678 provider name and the tracepoint name. The corresponding tracepoint
2679 definition defines the other parameters.
2681 .`tracepoint()` usage.
2683 The following <<defining-tracepoints,tracepoint definition>> defines a
2684 tracepoint which takes two input arguments and has two output event
2688 .Tracepoint provider header file.
2690 #include "my-custom-structure.h"
2697 const char*, cmd_name
2700 ctf_string(cmd_name, cmd_name)
2701 ctf_integer(int, number_of_args, argc)
2706 You can refer to this tracepoint definition with the `tracepoint()`
2707 macro in your application's source code like this:
2710 .Application's source file.
2714 int main(int argc, char* argv[])
2716 tracepoint(my_provider, my_tracepoint, argc, argv[0]);
2722 Note how the application's source code includes
2723 the tracepoint provider header file containing the tracepoint
2724 definitions to use, path:{tp.h}.
2727 .`tracepoint()` usage with a complex tracepoint definition.
2729 Consider this complex tracepoint definition, where multiple event
2730 fields refer to the same input arguments in their argument expression
2734 .Tracepoint provider header file.
2736 /* For `struct stat` */
2737 #include <sys/types.h>
2738 #include <sys/stat.h>
2750 ctf_integer(int, my_constant_field, 23 + 17)
2751 ctf_integer(int, my_int_arg_field, my_int_arg)
2752 ctf_integer(int, my_int_arg_field2, my_int_arg * my_int_arg)
2753 ctf_integer(int, sum4_field, my_str_arg[0] + my_str_arg[1] +
2754 my_str_arg[2] + my_str_arg[3])
2755 ctf_string(my_str_arg_field, my_str_arg)
2756 ctf_integer_hex(off_t, size_field, st->st_size)
2757 ctf_float(double, size_dbl_field, (double) st->st_size)
2758 ctf_sequence_text(char, half_my_str_arg_field, my_str_arg,
2759 size_t, strlen(my_str_arg) / 2)
2764 You can refer to this tracepoint definition with the `tracepoint()`
2765 macro in your application's source code like this:
2768 .Application's source file.
2770 #define TRACEPOINT_DEFINE
2777 stat("/etc/fstab", &s);
2778 tracepoint(my_provider, my_tracepoint, 23, "Hello, World!", &s);
2784 If you look at the event record that LTTng writes when tracing this
2785 program, assuming the file size of path:{/etc/fstab} is 301{nbsp}bytes,
2786 it should look like this:
2788 .Event record fields
2790 |Field's name |Field's value
2791 |`my_constant_field` |40
2792 |`my_int_arg_field` |23
2793 |`my_int_arg_field2` |529
2795 |`my_str_arg_field` |`Hello, World!`
2796 |`size_field` |0x12d
2797 |`size_dbl_field` |301.0
2798 |`half_my_str_arg_field` |`Hello,`
2802 Sometimes, the arguments you pass to `tracepoint()` are expensive to
2803 compute--they use the call stack, for example. To avoid this
2804 computation when the tracepoint is disabled, you can use the
2805 `tracepoint_enabled()` and `do_tracepoint()` macros.
2807 The syntax of the `tracepoint_enabled()` and `do_tracepoint()` macros
2811 .`tracepoint_enabled()` and `do_tracepoint()` macros syntax.
2813 tracepoint_enabled(provider_name, tracepoint_name)
2814 do_tracepoint(provider_name, tracepoint_name, ...)
2819 * `provider_name` with the tracepoint provider name.
2820 * `tracepoint_name` with the tracepoint name.
2822 `tracepoint_enabled()` returns a non-zero value if the tracepoint named
2823 `tracepoint_name` from the provider named `provider_name` is enabled
2826 `do_tracepoint()` is like `tracepoint()`, except that it doesn't check
2827 if the tracepoint is enabled. Using `tracepoint()` with
2828 `tracepoint_enabled()` is dangerous since `tracepoint()` also contains
2829 the `tracepoint_enabled()` check, thus a race condition is
2830 possible in this situation:
2833 .Possible race condition when using `tracepoint_enabled()` with `tracepoint()`.
2835 if (tracepoint_enabled(my_provider, my_tracepoint)) {
2836 stuff = prepare_stuff();
2839 tracepoint(my_provider, my_tracepoint, stuff);
2842 If the tracepoint is enabled after the condition, then `stuff` is not
2843 prepared: the emitted event will either contain wrong data, or the whole
2844 application could crash (segmentation fault, for example).
2846 NOTE: Neither `tracepoint_enabled()` nor `do_tracepoint()` have an
2847 `STAP_PROBEV()` call. If you need it, you must emit
2851 [[building-tracepoint-providers-and-user-application]]
2852 ==== Build and link a tracepoint provider package and an application
2854 Once you have one or more <<tpp-header,tracepoint provider header
2855 files>> and a <<tpp-source,tracepoint provider package source file>>,
2856 you can create the tracepoint provider package by compiling its source
2857 file. From here, multiple build and run scenarios are possible. The
2858 following table shows common application and library configurations
2859 along with the required command lines to achieve them.
2861 In the following diagrams, we use the following file names:
2864 Executable application.
2867 Application's object file.
2870 Tracepoint provider package object file.
2873 Tracepoint provider package archive file.
2876 Tracepoint provider package shared object file.
2879 User library object file.
2882 User library shared object file.
2884 We use the following symbols in the diagrams of table below:
2887 .Symbols used in the build scenario diagrams.
2888 image::ust-sit-symbols.png[]
2890 We assume that path:{.} is part of the env:LD_LIBRARY_PATH environment
2891 variable in the following instructions.
2893 [role="growable ust-scenarios",cols="asciidoc,asciidoc"]
2894 .Common tracepoint provider package scenarios.
2896 |Scenario |Instructions
2899 The instrumented application is statically linked with
2900 the tracepoint provider package object.
2902 image::ust-sit+app-linked-with-tp-o+app-instrumented.png[]
2905 include::../common/ust-sit-step-tp-o.txt[]
2907 To build the instrumented application:
2909 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2914 #define TRACEPOINT_DEFINE
2918 . Compile the application source file:
2927 . Build the application:
2932 $ gcc -o app app.o tpp.o -llttng-ust -ldl
2936 To run the instrumented application:
2938 * Start the application:
2948 The instrumented application is statically linked with the
2949 tracepoint provider package archive file.
2951 image::ust-sit+app-linked-with-tp-a+app-instrumented.png[]
2954 To create the tracepoint provider package archive file:
2956 . Compile the <<tpp-source,tracepoint provider package source file>>:
2965 . Create the tracepoint provider package archive file:
2970 $ ar rcs tpp.a tpp.o
2974 To build the instrumented application:
2976 . In path:{app.c}, before including path:{tpp.h}, add the following line:
2981 #define TRACEPOINT_DEFINE
2985 . Compile the application source file:
2994 . Build the application:
2999 $ gcc -o app app.o tpp.a -llttng-ust -ldl
3003 To run the instrumented application:
3005 * Start the application:
3015 The instrumented application is linked with the tracepoint provider
3016 package shared object.
3018 image::ust-sit+app-linked-with-tp-so+app-instrumented.png[]
3021 include::../common/ust-sit-step-tp-so.txt[]
3023 To build the instrumented application:
3025 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3030 #define TRACEPOINT_DEFINE
3034 . Compile the application source file:
3043 . Build the application:
3048 $ gcc -o app app.o -ldl -L. -ltpp
3052 To run the instrumented application:
3054 * Start the application:
3064 The tracepoint provider package shared object is preloaded before the
3065 instrumented application starts.
3067 image::ust-sit+tp-so-preloaded+app-instrumented.png[]
3070 include::../common/ust-sit-step-tp-so.txt[]
3072 To build the instrumented application:
3074 . In path:{app.c}, before including path:{tpp.h}, add the
3080 #define TRACEPOINT_DEFINE
3081 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3085 . Compile the application source file:
3094 . Build the application:
3099 $ gcc -o app app.o -ldl
3103 To run the instrumented application with tracing support:
3105 * Preload the tracepoint provider package shared object and
3106 start the application:
3111 $ LD_PRELOAD=./libtpp.so ./app
3115 To run the instrumented application without tracing support:
3117 * Start the application:
3127 The instrumented application dynamically loads the tracepoint provider
3128 package shared object.
3130 See the <<dlclose-warning,warning about `dlclose()`>>.
3132 image::ust-sit+app-dlopens-tp-so+app-instrumented.png[]
3135 include::../common/ust-sit-step-tp-so.txt[]
3137 To build the instrumented application:
3139 . In path:{app.c}, before including path:{tpp.h}, add the
3145 #define TRACEPOINT_DEFINE
3146 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3150 . Compile the application source file:
3159 . Build the application:
3164 $ gcc -o app app.o -ldl
3168 To run the instrumented application:
3170 * Start the application:
3180 The application is linked with the instrumented user library.
3182 The instrumented user library is statically linked with the tracepoint
3183 provider package object file.
3185 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-o+lib-instrumented.png[]
3188 include::../common/ust-sit-step-tp-o-fpic.txt[]
3190 To build the instrumented user library:
3192 . In path:{emon.c}, before including path:{tpp.h}, add the
3198 #define TRACEPOINT_DEFINE
3202 . Compile the user library source file:
3207 $ gcc -I. -fpic -c emon.c
3211 . Build the user library shared object:
3216 $ gcc -shared -o libemon.so emon.o tpp.o -llttng-ust -ldl
3220 To build the application:
3222 . Compile the application source file:
3231 . Build the application:
3236 $ gcc -o app app.o -L. -lemon
3240 To run the application:
3242 * Start the application:
3252 The application is linked with the instrumented user library.
3254 The instrumented user library is linked with the tracepoint provider
3255 package shared object.
3257 image::ust-sit+app-linked-with-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3260 include::../common/ust-sit-step-tp-so.txt[]
3262 To build the instrumented user library:
3264 . In path:{emon.c}, before including path:{tpp.h}, add the
3270 #define TRACEPOINT_DEFINE
3274 . Compile the user library source file:
3279 $ gcc -I. -fpic -c emon.c
3283 . Build the user library shared object:
3288 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3292 To build the application:
3294 . Compile the application source file:
3303 . Build the application:
3308 $ gcc -o app app.o -L. -lemon
3312 To run the application:
3314 * Start the application:
3324 The tracepoint provider package shared object is preloaded before the
3327 The application is linked with the instrumented user library.
3329 image::ust-sit+tp-so-preloaded+app-linked-with-lib+lib-instrumented.png[]
3332 include::../common/ust-sit-step-tp-so.txt[]
3334 To build the instrumented user library:
3336 . In path:{emon.c}, before including path:{tpp.h}, add the
3342 #define TRACEPOINT_DEFINE
3343 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3347 . Compile the user library source file:
3352 $ gcc -I. -fpic -c emon.c
3356 . Build the user library shared object:
3361 $ gcc -shared -o libemon.so emon.o -ldl
3365 To build the application:
3367 . Compile the application source file:
3376 . Build the application:
3381 $ gcc -o app app.o -L. -lemon
3385 To run the application with tracing support:
3387 * Preload the tracepoint provider package shared object and
3388 start the application:
3393 $ LD_PRELOAD=./libtpp.so ./app
3397 To run the application without tracing support:
3399 * Start the application:
3409 The application is linked with the instrumented user library.
3411 The instrumented user library dynamically loads the tracepoint provider
3412 package shared object.
3414 See the <<dlclose-warning,warning about `dlclose()`>>.
3416 image::ust-sit+app-linked-with-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3419 include::../common/ust-sit-step-tp-so.txt[]
3421 To build the instrumented user library:
3423 . In path:{emon.c}, before including path:{tpp.h}, add the
3429 #define TRACEPOINT_DEFINE
3430 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3434 . Compile the user library source file:
3439 $ gcc -I. -fpic -c emon.c
3443 . Build the user library shared object:
3448 $ gcc -shared -o libemon.so emon.o -ldl
3452 To build the application:
3454 . Compile the application source file:
3463 . Build the application:
3468 $ gcc -o app app.o -L. -lemon
3472 To run the application:
3474 * Start the application:
3484 The application dynamically loads the instrumented user library.
3486 The instrumented user library is linked with the tracepoint provider
3487 package shared object.
3489 See the <<dlclose-warning,warning about `dlclose()`>>.
3491 image::ust-sit+app-dlopens-lib+lib-linked-with-tp-so+lib-instrumented.png[]
3494 include::../common/ust-sit-step-tp-so.txt[]
3496 To build the instrumented user library:
3498 . In path:{emon.c}, before including path:{tpp.h}, add the
3504 #define TRACEPOINT_DEFINE
3508 . Compile the user library source file:
3513 $ gcc -I. -fpic -c emon.c
3517 . Build the user library shared object:
3522 $ gcc -shared -o libemon.so emon.o -ldl -L. -ltpp
3526 To build the application:
3528 . Compile the application source file:
3537 . Build the application:
3542 $ gcc -o app app.o -ldl -L. -lemon
3546 To run the application:
3548 * Start the application:
3558 The application dynamically loads the instrumented user library.
3560 The instrumented user library dynamically loads the tracepoint provider
3561 package shared object.
3563 See the <<dlclose-warning,warning about `dlclose()`>>.
3565 image::ust-sit+app-dlopens-lib+lib-dlopens-tp-so+lib-instrumented.png[]
3568 include::../common/ust-sit-step-tp-so.txt[]
3570 To build the instrumented user library:
3572 . In path:{emon.c}, before including path:{tpp.h}, add the
3578 #define TRACEPOINT_DEFINE
3579 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3583 . Compile the user library source file:
3588 $ gcc -I. -fpic -c emon.c
3592 . Build the user library shared object:
3597 $ gcc -shared -o libemon.so emon.o -ldl
3601 To build the application:
3603 . Compile the application source file:
3612 . Build the application:
3617 $ gcc -o app app.o -ldl -L. -lemon
3621 To run the application:
3623 * Start the application:
3633 The tracepoint provider package shared object is preloaded before the
3636 The application dynamically loads the instrumented user library.
3638 image::ust-sit+tp-so-preloaded+app-dlopens-lib+lib-instrumented.png[]
3641 include::../common/ust-sit-step-tp-so.txt[]
3643 To build the instrumented user library:
3645 . In path:{emon.c}, before including path:{tpp.h}, add the
3651 #define TRACEPOINT_DEFINE
3652 #define TRACEPOINT_PROBE_DYNAMIC_LINKAGE
3656 . Compile the user library source file:
3661 $ gcc -I. -fpic -c emon.c
3665 . Build the user library shared object:
3670 $ gcc -shared -o libemon.so emon.o -ldl
3674 To build the application:
3676 . Compile the application source file:
3685 . Build the application:
3690 $ gcc -o app app.o -L. -lemon
3694 To run the application with tracing support:
3696 * Preload the tracepoint provider package shared object and
3697 start the application:
3702 $ LD_PRELOAD=./libtpp.so ./app
3706 To run the application without tracing support:
3708 * Start the application:
3718 The application is statically linked with the tracepoint provider
3719 package object file.
3721 The application is linked with the instrumented user library.
3723 image::ust-sit+app-linked-with-tp-o+app-linked-with-lib+lib-instrumented.png[]
3726 include::../common/ust-sit-step-tp-o.txt[]
3728 To build the instrumented user library:
3730 . In path:{emon.c}, before including path:{tpp.h}, add the
3736 #define TRACEPOINT_DEFINE
3740 . Compile the user library source file:
3745 $ gcc -I. -fpic -c emon.c
3749 . Build the user library shared object:
3754 $ gcc -shared -o libemon.so emon.o
3758 To build the application:
3760 . Compile the application source file:
3769 . Build the application:
3774 $ gcc -o app app.o tpp.o -llttng-ust -ldl -L. -lemon
3778 To run the instrumented application:
3780 * Start the application:
3790 The application is statically linked with the tracepoint provider
3791 package object file.
3793 The application dynamically loads the instrumented user library.
3795 image::ust-sit+app-linked-with-tp-o+app-dlopens-lib+lib-instrumented.png[]
3798 include::../common/ust-sit-step-tp-o.txt[]
3800 To build the application:
3802 . In path:{app.c}, before including path:{tpp.h}, add the following line:
3807 #define TRACEPOINT_DEFINE
3811 . Compile the application source file:
3820 . Build the application:
3825 $ gcc -Wl,--export-dynamic -o app app.o tpp.o \
3830 The `--export-dynamic` option passed to the linker is necessary for the
3831 dynamically loaded library to ``see'' the tracepoint symbols defined in
3834 To build the instrumented user library:
3836 . Compile the user library source file:
3841 $ gcc -I. -fpic -c emon.c
3845 . Build the user library shared object:
3850 $ gcc -shared -o libemon.so emon.o
3854 To run the application:
3856 * Start the application:
3868 .Do not use man:dlclose(3) on a tracepoint provider package
3870 Never use man:dlclose(3) on any shared object which:
3872 * Is linked with, statically or dynamically, a tracepoint provider
3874 * Calls man:dlopen(3) itself to dynamically open a tracepoint provider
3875 package shared object.
3877 This is currently considered **unsafe** due to a lack of reference
3878 counting from LTTng-UST to the shared object.
3880 A known workaround (available since glibc 2.2) is to use the
3881 `RTLD_NODELETE` flag when calling man:dlopen(3) initially. This has the
3882 effect of not unloading the loaded shared object, even if man:dlclose(3)
3885 You can also preload the tracepoint provider package shared object with
3886 the env:LD_PRELOAD environment variable to overcome this limitation.
3890 [[using-lttng-ust-with-daemons]]
3891 ===== Use noch:{LTTng-UST} with daemons
3893 If your instrumented application calls man:fork(2), man:clone(2),
3894 or BSD's man:rfork(2), without a following man:exec(3)-family
3895 system call, you must preload the path:{liblttng-ust-fork.so} shared
3896 object when you start the application.
3900 $ LD_PRELOAD=liblttng-ust-fork.so ./my-app
3903 If your tracepoint provider package is
3904 a shared library which you also preload, you must put both
3905 shared objects in env:LD_PRELOAD:
3909 $ LD_PRELOAD=liblttng-ust-fork.so:/path/to/tp.so ./my-app
3915 ===== Use noch:{LTTng-UST} with applications which close file descriptors that don't belong to them
3917 If your instrumented application closes one or more file descriptors
3918 which it did not open itself, you must preload the
3919 path:{liblttng-ust-fd.so} shared object when you start the application:
3923 $ LD_PRELOAD=liblttng-ust-fd.so ./my-app
3926 Typical use cases include closing all the file descriptors after
3927 man:fork(2) or man:rfork(2) and buggy applications doing
3931 [[lttng-ust-pkg-config]]
3932 ===== Use noch:{pkg-config}
3934 On some distributions, LTTng-UST ships with a
3935 https://www.freedesktop.org/wiki/Software/pkg-config/[pkg-config]
3936 metadata file. If this is your case, then you can use cmd:pkg-config to
3937 build an application on the command line:
3941 $ gcc -o my-app my-app.o tp.o $(pkg-config --cflags --libs lttng-ust)
3945 [[instrumenting-32-bit-app-on-64-bit-system]]
3946 ===== [[advanced-instrumenting-techniques]]Build a 32-bit instrumented application for a 64-bit target system
3948 In order to trace a 32-bit application running on a 64-bit system,
3949 LTTng must use a dedicated 32-bit
3950 <<lttng-consumerd,consumer daemon>>.
3952 The following steps show how to build and install a 32-bit consumer
3953 daemon, which is _not_ part of the default 64-bit LTTng build, how to
3954 build and install the 32-bit LTTng-UST libraries, and how to build and
3955 link an instrumented 32-bit application in that context.
3957 To build a 32-bit instrumented application for a 64-bit target system,
3958 assuming you have a fresh target system with no installed Userspace RCU
3961 . Download, build, and install a 32-bit version of Userspace RCU:
3966 $ cd $(mktemp -d) &&
3967 wget http://lttng.org/files/urcu/userspace-rcu-latest-0.9.tar.bz2 &&
3968 tar -xf userspace-rcu-latest-0.9.tar.bz2 &&
3969 cd userspace-rcu-0.9.* &&
3970 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 &&
3972 sudo make install &&
3977 . Using your distribution's package manager, or from source, install
3978 the following 32-bit versions of the following dependencies of
3979 LTTng-tools and LTTng-UST:
3982 * https://sourceforge.net/projects/libuuid/[libuuid]
3983 * http://directory.fsf.org/wiki/Popt[popt]
3984 * http://www.xmlsoft.org/[libxml2]
3987 . Download, build, and install a 32-bit version of the latest
3988 LTTng-UST{nbsp}{revision}:
3993 $ cd $(mktemp -d) &&
3994 wget http://lttng.org/files/lttng-ust/lttng-ust-latest-2.10.tar.bz2 &&
3995 tar -xf lttng-ust-latest-2.10.tar.bz2 &&
3996 cd lttng-ust-2.10.* &&
3997 ./configure --libdir=/usr/local/lib32 \
3998 CFLAGS=-m32 CXXFLAGS=-m32 \
3999 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' &&
4001 sudo make install &&
4008 Depending on your distribution,
4009 32-bit libraries could be installed at a different location than
4010 `/usr/lib32`. For example, Debian is known to install
4011 some 32-bit libraries in `/usr/lib/i386-linux-gnu`.
4013 In this case, make sure to set `LDFLAGS` to all the
4014 relevant 32-bit library paths, for example:
4018 $ LDFLAGS='-L/usr/lib/i386-linux-gnu -L/usr/lib32'
4022 . Download the latest LTTng-tools{nbsp}{revision}, build, and install
4023 the 32-bit consumer daemon:
4028 $ cd $(mktemp -d) &&
4029 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
4030 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
4031 cd lttng-tools-2.10.* &&
4032 ./configure --libdir=/usr/local/lib32 CFLAGS=-m32 CXXFLAGS=-m32 \
4033 LDFLAGS='-L/usr/local/lib32 -L/usr/lib32' \
4034 --disable-bin-lttng --disable-bin-lttng-crash \
4035 --disable-bin-lttng-relayd --disable-bin-lttng-sessiond &&
4037 cd src/bin/lttng-consumerd &&
4038 sudo make install &&
4043 . From your distribution or from source,
4044 <<installing-lttng,install>> the 64-bit versions of
4045 LTTng-UST and Userspace RCU.
4046 . Download, build, and install the 64-bit version of the
4047 latest LTTng-tools{nbsp}{revision}:
4052 $ cd $(mktemp -d) &&
4053 wget http://lttng.org/files/lttng-tools/lttng-tools-latest-2.10.tar.bz2 &&
4054 tar -xf lttng-tools-latest-2.10.tar.bz2 &&
4055 cd lttng-tools-2.10.* &&
4056 ./configure --with-consumerd32-libdir=/usr/local/lib32 \
4057 --with-consumerd32-bin=/usr/local/lib32/lttng/libexec/lttng-consumerd &&
4059 sudo make install &&
4064 . Pass the following options to man:gcc(1), man:g++(1), or man:clang(1)
4065 when linking your 32-bit application:
4068 -m32 -L/usr/lib32 -L/usr/local/lib32 \
4069 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32
4072 For example, let's rebuild the quick start example in
4073 <<tracing-your-own-user-application,Trace a user application>> as an
4074 instrumented 32-bit application:
4079 $ gcc -m32 -c -I. hello-tp.c
4080 $ gcc -m32 -c hello.c
4081 $ gcc -m32 -o hello hello.o hello-tp.o \
4082 -L/usr/lib32 -L/usr/local/lib32 \
4083 -Wl,-rpath,/usr/lib32,-rpath,/usr/local/lib32 \
4088 No special action is required to execute the 32-bit application and
4089 to trace it: use the command-line man:lttng(1) tool as usual.
4096 man:tracef(3) is a small LTTng-UST API designed for quick,
4097 man:printf(3)-like instrumentation without the burden of
4098 <<tracepoint-provider,creating>> and
4099 <<building-tracepoint-providers-and-user-application,building>>
4100 a tracepoint provider package.
4102 To use `tracef()` in your application:
4104 . In the C or C++ source files where you need to use `tracef()`,
4105 include `<lttng/tracef.h>`:
4110 #include <lttng/tracef.h>
4114 . In the application's source code, use `tracef()` like you would use
4122 tracef("my message: %d (%s)", my_integer, my_string);
4128 . Link your application with `liblttng-ust`:
4133 $ gcc -o app app.c -llttng-ust
4137 To trace the events that `tracef()` calls emit:
4139 * <<enabling-disabling-events,Create an event rule>> which matches the
4140 `lttng_ust_tracef:*` event name:
4145 $ lttng enable-event --userspace 'lttng_ust_tracef:*'
4150 .Limitations of `tracef()`
4152 The `tracef()` utility function was developed to make user space tracing
4153 super simple, albeit with notable disadvantages compared to
4154 <<defining-tracepoints,user-defined tracepoints>>:
4156 * All the emitted events have the same tracepoint provider and
4157 tracepoint names, respectively `lttng_ust_tracef` and `event`.
4158 * There is no static type checking.
4159 * The only event record field you actually get, named `msg`, is a string
4160 potentially containing the values you passed to `tracef()`
4161 using your own format string. This also means that you cannot filter
4162 events with a custom expression at run time because there are no
4164 * Since `tracef()` uses the C standard library's man:vasprintf(3)
4165 function behind the scenes to format the strings at run time, its
4166 expected performance is lower than with user-defined tracepoints,
4167 which do not require a conversion to a string.
4169 Taking this into consideration, `tracef()` is useful for some quick
4170 prototyping and debugging, but you should not consider it for any
4171 permanent and serious applicative instrumentation.
4177 ==== Use `tracelog()`
4179 The man:tracelog(3) API is very similar to <<tracef,`tracef()`>>, with
4180 the difference that it accepts an additional log level parameter.
4182 The goal of `tracelog()` is to ease the migration from logging to
4185 To use `tracelog()` in your application:
4187 . In the C or C++ source files where you need to use `tracelog()`,
4188 include `<lttng/tracelog.h>`:
4193 #include <lttng/tracelog.h>
4197 . In the application's source code, use `tracelog()` like you would use
4198 man:printf(3), except for the first parameter which is the log
4206 tracelog(TRACE_WARNING, "my message: %d (%s)",
4207 my_integer, my_string);
4213 See man:lttng-ust(3) for a list of available log level names.
4215 . Link your application with `liblttng-ust`:
4220 $ gcc -o app app.c -llttng-ust
4224 To trace the events that `tracelog()` calls emit with a log level
4225 _as severe as_ a specific log level:
4227 * <<enabling-disabling-events,Create an event rule>> which matches the
4228 `lttng_ust_tracelog:*` event name and a minimum level
4234 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
4235 --loglevel=TRACE_WARNING
4239 To trace the events that `tracelog()` calls emit with a
4240 _specific log level_:
4242 * Create an event rule which matches the `lttng_ust_tracelog:*`
4243 event name and a specific log level:
4248 $ lttng enable-event --userspace 'lttng_ust_tracelog:*'
4249 --loglevel-only=TRACE_INFO
4254 [[prebuilt-ust-helpers]]
4255 === Prebuilt user space tracing helpers
4257 The LTTng-UST package provides a few helpers in the form or preloadable
4258 shared objects which automatically instrument system functions and
4261 The helper shared objects are normally found in dir:{/usr/lib}. If you
4262 built LTTng-UST <<building-from-source,from source>>, they are probably
4263 located in dir:{/usr/local/lib}.
4265 The installed user space tracing helpers in LTTng-UST{nbsp}{revision}
4268 path:{liblttng-ust-libc-wrapper.so}::
4269 path:{liblttng-ust-pthread-wrapper.so}::
4270 <<liblttng-ust-libc-pthread-wrapper,C{nbsp}standard library
4271 memory and POSIX threads function tracing>>.
4273 path:{liblttng-ust-cyg-profile.so}::
4274 path:{liblttng-ust-cyg-profile-fast.so}::
4275 <<liblttng-ust-cyg-profile,Function entry and exit tracing>>.
4277 path:{liblttng-ust-dl.so}::
4278 <<liblttng-ust-dl,Dynamic linker tracing>>.
4280 To use a user space tracing helper with any user application:
4282 * Preload the helper shared object when you start the application:
4287 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so my-app
4291 You can preload more than one helper:
4296 $ LD_PRELOAD=liblttng-ust-libc-wrapper.so:liblttng-ust-dl.so my-app
4302 [[liblttng-ust-libc-pthread-wrapper]]
4303 ==== Instrument C standard library memory and POSIX threads functions
4305 The path:{liblttng-ust-libc-wrapper.so} and
4306 path:{liblttng-ust-pthread-wrapper.so} helpers
4307 add instrumentation to some C standard library and POSIX
4311 .Functions instrumented by preloading path:{liblttng-ust-libc-wrapper.so}.
4313 |TP provider name |TP name |Instrumented function
4315 .6+|`lttng_ust_libc` |`malloc` |man:malloc(3)
4316 |`calloc` |man:calloc(3)
4317 |`realloc` |man:realloc(3)
4318 |`free` |man:free(3)
4319 |`memalign` |man:memalign(3)
4320 |`posix_memalign` |man:posix_memalign(3)
4324 .Functions instrumented by preloading path:{liblttng-ust-pthread-wrapper.so}.
4326 |TP provider name |TP name |Instrumented function
4328 .4+|`lttng_ust_pthread` |`pthread_mutex_lock_req` |man:pthread_mutex_lock(3p) (request time)
4329 |`pthread_mutex_lock_acq` |man:pthread_mutex_lock(3p) (acquire time)
4330 |`pthread_mutex_trylock` |man:pthread_mutex_trylock(3p)
4331 |`pthread_mutex_unlock` |man:pthread_mutex_unlock(3p)
4334 When you preload the shared object, it replaces the functions listed
4335 in the previous tables by wrappers which contain tracepoints and call
4336 the replaced functions.
4339 [[liblttng-ust-cyg-profile]]
4340 ==== Instrument function entry and exit
4342 The path:{liblttng-ust-cyg-profile*.so} helpers can add instrumentation
4343 to the entry and exit points of functions.
4345 man:gcc(1) and man:clang(1) have an option named
4346 https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html[`-finstrument-functions`]
4347 which generates instrumentation calls for entry and exit to functions.
4348 The LTTng-UST function tracing helpers,
4349 path:{liblttng-ust-cyg-profile.so} and
4350 path:{liblttng-ust-cyg-profile-fast.so}, take advantage of this feature
4351 to add tracepoints to the two generated functions (which contain
4352 `cyg_profile` in their names, hence the helper's name).
4354 To use the LTTng-UST function tracing helper, the source files to
4355 instrument must be built using the `-finstrument-functions` compiler
4358 There are two versions of the LTTng-UST function tracing helper:
4360 * **path:{liblttng-ust-cyg-profile-fast.so}** is a lightweight variant
4361 that you should only use when it can be _guaranteed_ that the
4362 complete event stream is recorded without any lost event record.
4363 Any kind of duplicate information is left out.
4365 Assuming no event record is lost, having only the function addresses on
4366 entry is enough to create a call graph, since an event record always
4367 contains the ID of the CPU that generated it.
4369 You can use a tool like man:addr2line(1) to convert function addresses
4370 back to source file names and line numbers.
4372 * **path:{liblttng-ust-cyg-profile.so}** is a more robust variant
4373 which also works in use cases where event records might get discarded or
4374 not recorded from application startup.
4375 In these cases, the trace analyzer needs more information to be
4376 able to reconstruct the program flow.
4378 See man:lttng-ust-cyg-profile(3) to learn more about the instrumentation
4379 points of this helper.
4381 All the tracepoints that this helper provides have the
4382 log level `TRACE_DEBUG_FUNCTION` (see man:lttng-ust(3)).
4384 TIP: It's sometimes a good idea to limit the number of source files that
4385 you compile with the `-finstrument-functions` option to prevent LTTng
4386 from writing an excessive amount of trace data at run time. When using
4387 man:gcc(1), you can use the
4388 `-finstrument-functions-exclude-function-list` option to avoid
4389 instrument entries and exits of specific function names.
4394 ==== Instrument the dynamic linker
4396 The path:{liblttng-ust-dl.so} helper adds instrumentation to the
4397 man:dlopen(3) and man:dlclose(3) function calls.
4399 See man:lttng-ust-dl(3) to learn more about the instrumentation points
4404 [[java-application]]
4405 === User space Java agent
4407 You can instrument any Java application which uses one of the following
4410 * The https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[**`java.util.logging`**]
4411 (JUL) core logging facilities.
4412 * http://logging.apache.org/log4j/1.2/[**Apache log4j 1.2**], since
4413 LTTng 2.6. Note that Apache Log4j{nbsp}2 is not supported.
4416 .LTTng-UST Java agent imported by a Java application.
4417 image::java-app.png[]
4419 Note that the methods described below are new in LTTng{nbsp}{revision}.
4420 Previous LTTng versions use another technique.
4422 NOTE: We use http://openjdk.java.net/[OpenJDK]{nbsp}8 for development
4423 and https://ci.lttng.org/[continuous integration], thus this version is
4424 directly supported. However, the LTTng-UST Java agent is also tested
4425 with OpenJDK{nbsp}7.
4430 ==== Use the LTTng-UST Java agent for `java.util.logging`
4432 To use the LTTng-UST Java agent in a Java application which uses
4433 `java.util.logging` (JUL):
4435 . In the Java application's source code, import the LTTng-UST
4436 log handler package for `java.util.logging`:
4441 import org.lttng.ust.agent.jul.LttngLogHandler;
4445 . Create an LTTng-UST JUL log handler:
4450 Handler lttngUstLogHandler = new LttngLogHandler();
4454 . Add this handler to the JUL loggers which should emit LTTng events:
4459 Logger myLogger = Logger.getLogger("some-logger");
4461 myLogger.addHandler(lttngUstLogHandler);
4465 . Use `java.util.logging` log statements and configuration as usual.
4466 The loggers with an attached LTTng-UST log handler can emit
4469 . Before exiting the application, remove the LTTng-UST log handler from
4470 the loggers attached to it and call its `close()` method:
4475 myLogger.removeHandler(lttngUstLogHandler);
4476 lttngUstLogHandler.close();
4480 This is not strictly necessary, but it is recommended for a clean
4481 disposal of the handler's resources.
4483 . Include the LTTng-UST Java agent's common and JUL-specific JAR files,
4484 path:{lttng-ust-agent-common.jar} and path:{lttng-ust-agent-jul.jar},
4486 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4487 path] when you build the Java application.
4489 The JAR files are typically located in dir:{/usr/share/java}.
4491 IMPORTANT: The LTTng-UST Java agent must be
4492 <<installing-lttng,installed>> for the logging framework your
4495 .Use the LTTng-UST Java agent for `java.util.logging`.
4500 import java.io.IOException;
4501 import java.util.logging.Handler;
4502 import java.util.logging.Logger;
4503 import org.lttng.ust.agent.jul.LttngLogHandler;
4507 private static final int answer = 42;
4509 public static void main(String[] argv) throws Exception
4512 Logger logger = Logger.getLogger("jello");
4514 // Create an LTTng-UST log handler
4515 Handler lttngUstLogHandler = new LttngLogHandler();
4517 // Add the LTTng-UST log handler to our logger
4518 logger.addHandler(lttngUstLogHandler);
4521 logger.info("some info");
4522 logger.warning("some warning");
4524 logger.finer("finer information; the answer is " + answer);
4526 logger.severe("error!");
4528 // Not mandatory, but cleaner
4529 logger.removeHandler(lttngUstLogHandler);
4530 lttngUstLogHandler.close();
4539 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4542 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4543 <<enabling-disabling-events,create an event rule>> matching the
4544 `jello` JUL logger, and <<basic-tracing-session-control,start tracing>>:
4549 $ lttng enable-event --jul jello
4553 Run the compiled class:
4557 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4560 <<basic-tracing-session-control,Stop tracing>> and inspect the
4570 In the resulting trace, an <<event,event record>> generated by a Java
4571 application using `java.util.logging` is named `lttng_jul:event` and
4572 has the following fields:
4575 Log record's message.
4581 Name of the class in which the log statement was executed.
4584 Name of the method in which the log statement was executed.
4587 Logging time (timestamp in milliseconds).
4590 Log level integer value.
4593 ID of the thread in which the log statement was executed.
4595 You can use the opt:lttng-enable-event(1):--loglevel or
4596 opt:lttng-enable-event(1):--loglevel-only option of the
4597 man:lttng-enable-event(1) command to target a range of JUL log levels
4598 or a specific JUL log level.
4603 ==== Use the LTTng-UST Java agent for Apache log4j
4605 To use the LTTng-UST Java agent in a Java application which uses
4608 . In the Java application's source code, import the LTTng-UST
4609 log appender package for Apache log4j:
4614 import org.lttng.ust.agent.log4j.LttngLogAppender;
4618 . Create an LTTng-UST log4j log appender:
4623 Appender lttngUstLogAppender = new LttngLogAppender();
4627 . Add this appender to the log4j loggers which should emit LTTng events:
4632 Logger myLogger = Logger.getLogger("some-logger");
4634 myLogger.addAppender(lttngUstLogAppender);
4638 . Use Apache log4j log statements and configuration as usual. The
4639 loggers with an attached LTTng-UST log appender can emit LTTng events.
4641 . Before exiting the application, remove the LTTng-UST log appender from
4642 the loggers attached to it and call its `close()` method:
4647 myLogger.removeAppender(lttngUstLogAppender);
4648 lttngUstLogAppender.close();
4652 This is not strictly necessary, but it is recommended for a clean
4653 disposal of the appender's resources.
4655 . Include the LTTng-UST Java agent's common and log4j-specific JAR
4656 files, path:{lttng-ust-agent-common.jar} and
4657 path:{lttng-ust-agent-log4j.jar}, in the
4658 https://docs.oracle.com/javase/tutorial/essential/environment/paths.html[class
4659 path] when you build the Java application.
4661 The JAR files are typically located in dir:{/usr/share/java}.
4663 IMPORTANT: The LTTng-UST Java agent must be
4664 <<installing-lttng,installed>> for the logging framework your
4667 .Use the LTTng-UST Java agent for Apache log4j.
4672 import org.apache.log4j.Appender;
4673 import org.apache.log4j.Logger;
4674 import org.lttng.ust.agent.log4j.LttngLogAppender;
4678 private static final int answer = 42;
4680 public static void main(String[] argv) throws Exception
4683 Logger logger = Logger.getLogger("jello");
4685 // Create an LTTng-UST log appender
4686 Appender lttngUstLogAppender = new LttngLogAppender();
4688 // Add the LTTng-UST log appender to our logger
4689 logger.addAppender(lttngUstLogAppender);
4692 logger.info("some info");
4693 logger.warn("some warning");
4695 logger.debug("debug information; the answer is " + answer);
4697 logger.fatal("error!");
4699 // Not mandatory, but cleaner
4700 logger.removeAppender(lttngUstLogAppender);
4701 lttngUstLogAppender.close();
4707 Build this example (`$LOG4JPATH` is the path to the Apache log4j JAR
4712 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH Test.java
4715 <<creating-destroying-tracing-sessions,Create a tracing session>>,
4716 <<enabling-disabling-events,create an event rule>> matching the
4717 `jello` log4j logger, and <<basic-tracing-session-control,start tracing>>:
4722 $ lttng enable-event --log4j jello
4726 Run the compiled class:
4730 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-log4j.jar:$LOG4JPATH:. Test
4733 <<basic-tracing-session-control,Stop tracing>> and inspect the
4743 In the resulting trace, an <<event,event record>> generated by a Java
4744 application using log4j is named `lttng_log4j:event` and
4745 has the following fields:
4748 Log record's message.
4754 Name of the class in which the log statement was executed.
4757 Name of the method in which the log statement was executed.
4760 Name of the file in which the executed log statement is located.
4763 Line number at which the log statement was executed.
4769 Log level integer value.
4772 Name of the Java thread in which the log statement was executed.
4774 You can use the opt:lttng-enable-event(1):--loglevel or
4775 opt:lttng-enable-event(1):--loglevel-only option of the
4776 man:lttng-enable-event(1) command to target a range of Apache log4j log levels
4777 or a specific log4j log level.
4781 [[java-application-context]]
4782 ==== Provide application-specific context fields in a Java application
4784 A Java application-specific context field is a piece of state provided
4785 by the application which <<adding-context,you can add>>, using the
4786 man:lttng-add-context(1) command, to each <<event,event record>>
4787 produced by the log statements of this application.
4789 For example, a given object might have a current request ID variable.
4790 You can create a context information retriever for this object and
4791 assign a name to this current request ID. You can then, using the
4792 man:lttng-add-context(1) command, add this context field by name to
4793 the JUL or log4j <<channel,channel>>.
4795 To provide application-specific context fields in a Java application:
4797 . In the Java application's source code, import the LTTng-UST
4798 Java agent context classes and interfaces:
4803 import org.lttng.ust.agent.context.ContextInfoManager;
4804 import org.lttng.ust.agent.context.IContextInfoRetriever;
4808 . Create a context information retriever class, that is, a class which
4809 implements the `IContextInfoRetriever` interface:
4814 class MyContextInfoRetriever implements IContextInfoRetriever
4817 public Object retrieveContextInfo(String key)
4819 if (key.equals("intCtx")) {
4821 } else if (key.equals("strContext")) {
4822 return "context value!";
4831 This `retrieveContextInfo()` method is the only member of the
4832 `IContextInfoRetriever` interface. Its role is to return the current
4833 value of a state by name to create a context field. The names of the
4834 context fields and which state variables they return depends on your
4837 All primitive types and objects are supported as context fields.
4838 When `retrieveContextInfo()` returns an object, the context field
4839 serializer calls its `toString()` method to add a string field to
4840 event records. The method can also return `null`, which means that
4841 no context field is available for the required name.
4843 . Register an instance of your context information retriever class to
4844 the context information manager singleton:
4849 IContextInfoRetriever cir = new MyContextInfoRetriever();
4850 ContextInfoManager cim = ContextInfoManager.getInstance();
4851 cim.registerContextInfoRetriever("retrieverName", cir);
4855 . Before exiting the application, remove your context information
4856 retriever from the context information manager singleton:
4861 ContextInfoManager cim = ContextInfoManager.getInstance();
4862 cim.unregisterContextInfoRetriever("retrieverName");
4866 This is not strictly necessary, but it is recommended for a clean
4867 disposal of some manager's resources.
4869 . Build your Java application with LTTng-UST Java agent support as
4870 usual, following the procedure for either the <<jul,JUL>> or
4871 <<log4j,Apache log4j>> framework.
4874 .Provide application-specific context fields in a Java application.
4879 import java.util.logging.Handler;
4880 import java.util.logging.Logger;
4881 import org.lttng.ust.agent.jul.LttngLogHandler;
4882 import org.lttng.ust.agent.context.ContextInfoManager;
4883 import org.lttng.ust.agent.context.IContextInfoRetriever;
4887 // Our context information retriever class
4888 private static class MyContextInfoRetriever
4889 implements IContextInfoRetriever
4892 public Object retrieveContextInfo(String key) {
4893 if (key.equals("intCtx")) {
4895 } else if (key.equals("strContext")) {
4896 return "context value!";
4903 private static final int answer = 42;
4905 public static void main(String args[]) throws Exception
4907 // Get the context information manager instance
4908 ContextInfoManager cim = ContextInfoManager.getInstance();
4910 // Create and register our context information retriever
4911 IContextInfoRetriever cir = new MyContextInfoRetriever();
4912 cim.registerContextInfoRetriever("myRetriever", cir);
4915 Logger logger = Logger.getLogger("jello");
4917 // Create an LTTng-UST log handler
4918 Handler lttngUstLogHandler = new LttngLogHandler();
4920 // Add the LTTng-UST log handler to our logger
4921 logger.addHandler(lttngUstLogHandler);
4924 logger.info("some info");
4925 logger.warning("some warning");
4927 logger.finer("finer information; the answer is " + answer);
4929 logger.severe("error!");
4931 // Not mandatory, but cleaner
4932 logger.removeHandler(lttngUstLogHandler);
4933 lttngUstLogHandler.close();
4934 cim.unregisterContextInfoRetriever("myRetriever");
4943 $ javac -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar Test.java
4946 <<creating-destroying-tracing-sessions,Create a tracing session>>
4947 and <<enabling-disabling-events,create an event rule>> matching the
4953 $ lttng enable-event --jul jello
4956 <<adding-context,Add the application-specific context fields>> to the
4961 $ lttng add-context --jul --type='$app.myRetriever:intCtx'
4962 $ lttng add-context --jul --type='$app.myRetriever:strContext'
4965 <<basic-tracing-session-control,Start tracing>>:
4972 Run the compiled class:
4976 $ java -cp /usr/share/java/jarpath/lttng-ust-agent-common.jar:/usr/share/java/jarpath/lttng-ust-agent-jul.jar:. Test
4979 <<basic-tracing-session-control,Stop tracing>> and inspect the
4991 [[python-application]]
4992 === User space Python agent
4994 You can instrument a Python 2 or Python 3 application which uses the
4995 standard https://docs.python.org/3/library/logging.html[`logging`]
4998 Each log statement emits an LTTng event once the
4999 application module imports the
5000 <<lttng-ust-agents,LTTng-UST Python agent>> package.
5003 .A Python application importing the LTTng-UST Python agent.
5004 image::python-app.png[]
5006 To use the LTTng-UST Python agent:
5008 . In the Python application's source code, import the LTTng-UST Python
5018 The LTTng-UST Python agent automatically adds its logging handler to the
5019 root logger at import time.
5021 Any log statement that the application executes before this import does
5022 not emit an LTTng event.
5024 IMPORTANT: The LTTng-UST Python agent must be
5025 <<installing-lttng,installed>>.
5027 . Use log statements and logging configuration as usual.
5028 Since the LTTng-UST Python agent adds a handler to the _root_
5029 logger, you can trace any log statement from any logger.
5031 .Use the LTTng-UST Python agent.
5042 logging.basicConfig()
5043 logger = logging.getLogger('my-logger')
5046 logger.debug('debug message')
5047 logger.info('info message')
5048 logger.warn('warn message')
5049 logger.error('error message')
5050 logger.critical('critical message')
5054 if __name__ == '__main__':
5058 NOTE: `logging.basicConfig()`, which adds to the root logger a basic
5059 logging handler which prints to the standard error stream, is not
5060 strictly required for LTTng-UST tracing to work, but in versions of
5061 Python preceding 3.2, you could see a warning message which indicates
5062 that no handler exists for the logger `my-logger`.
5064 <<creating-destroying-tracing-sessions,Create a tracing session>>,
5065 <<enabling-disabling-events,create an event rule>> matching the
5066 `my-logger` Python logger, and <<basic-tracing-session-control,start
5072 $ lttng enable-event --python my-logger
5076 Run the Python script:
5083 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
5093 In the resulting trace, an <<event,event record>> generated by a Python
5094 application is named `lttng_python:event` and has the following fields:
5097 Logging time (string).
5100 Log record's message.
5106 Name of the function in which the log statement was executed.
5109 Line number at which the log statement was executed.
5112 Log level integer value.
5115 ID of the Python thread in which the log statement was executed.
5118 Name of the Python thread in which the log statement was executed.
5120 You can use the opt:lttng-enable-event(1):--loglevel or
5121 opt:lttng-enable-event(1):--loglevel-only option of the
5122 man:lttng-enable-event(1) command to target a range of Python log levels
5123 or a specific Python log level.
5125 When an application imports the LTTng-UST Python agent, the agent tries
5126 to register to a <<lttng-sessiond,session daemon>>. Note that you must
5127 <<start-sessiond,start the session daemon>> _before_ you run the Python
5128 application. If a session daemon is found, the agent tries to register
5129 to it during 5{nbsp}seconds, after which the application continues
5130 without LTTng tracing support. You can override this timeout value with
5131 the env:LTTNG_UST_PYTHON_REGISTER_TIMEOUT environment variable
5134 If the session daemon stops while a Python application with an imported
5135 LTTng-UST Python agent runs, the agent retries to connect and to
5136 register to a session daemon every 3{nbsp}seconds. You can override this
5137 delay with the env:LTTNG_UST_PYTHON_REGISTER_RETRY_DELAY environment
5142 [[proc-lttng-logger-abi]]
5145 The `lttng-tracer` Linux kernel module, part of
5146 <<lttng-modules,LTTng-modules>>, creates the special LTTng logger file
5147 path:{/proc/lttng-logger} when it's loaded. Any application can write
5148 text data to this file to emit an LTTng event.
5151 .An application writes to the LTTng logger file to emit an LTTng event.
5152 image::lttng-logger.png[]
5154 The LTTng logger is the quickest method--not the most efficient,
5155 however--to add instrumentation to an application. It is designed
5156 mostly to instrument shell scripts:
5160 $ echo "Some message, some $variable" > /proc/lttng-logger
5163 Any event that the LTTng logger emits is named `lttng_logger` and
5164 belongs to the Linux kernel <<domain,tracing domain>>. However, unlike
5165 other instrumentation points in the kernel tracing domain, **any Unix
5166 user** can <<enabling-disabling-events,create an event rule>> which
5167 matches its event name, not only the root user or users in the
5168 <<tracing-group,tracing group>>.
5170 To use the LTTng logger:
5172 * From any application, write text data to the path:{/proc/lttng-logger}
5175 The `msg` field of `lttng_logger` event records contains the
5178 NOTE: The maximum message length of an LTTng logger event is
5179 1024{nbsp}bytes. Writing more than this makes the LTTng logger emit more
5180 than one event to contain the remaining data.
5182 You should not use the LTTng logger to trace a user application which
5183 can be instrumented in a more efficient way, namely:
5185 * <<c-application,C and $$C++$$ applications>>.
5186 * <<java-application,Java applications>>.
5187 * <<python-application,Python applications>>.
5189 .Use the LTTng logger.
5194 echo 'Hello, World!' > /proc/lttng-logger
5196 df --human-readable --print-type / > /proc/lttng-logger
5199 <<creating-destroying-tracing-sessions,Create a tracing session>>,
5200 <<enabling-disabling-events,create an event rule>> matching the
5201 `lttng_logger` Linux kernel tracepoint, and
5202 <<basic-tracing-session-control,start tracing>>:
5207 $ lttng enable-event --kernel lttng_logger
5211 Run the Bash script:
5218 <<basic-tracing-session-control,Stop tracing>> and inspect the recorded
5229 [[instrumenting-linux-kernel]]
5230 === LTTng kernel tracepoints
5232 NOTE: This section shows how to _add_ instrumentation points to the
5233 Linux kernel. The kernel's subsystems are already thoroughly
5234 instrumented at strategic places for LTTng when you
5235 <<installing-lttng,install>> the <<lttng-modules,LTTng-modules>>
5239 There are two methods to instrument the Linux kernel:
5241 . <<linux-add-lttng-layer,Add an LTTng layer>> over an existing ftrace
5242 tracepoint which uses the `TRACE_EVENT()` API.
5244 Choose this if you want to instrumentation a Linux kernel tree with an
5245 instrumentation point compatible with ftrace, perf, and SystemTap.
5247 . Use an <<linux-lttng-tracepoint-event,LTTng-only approach>> to
5248 instrument an out-of-tree kernel module.
5250 Choose this if you don't need ftrace, perf, or SystemTap support.
5254 [[linux-add-lttng-layer]]
5255 ==== [[instrumenting-linux-kernel-itself]][[mainline-trace-event]][[lttng-adaptation-layer]]Add an LTTng layer to an existing ftrace tracepoint
5257 This section shows how to add an LTTng layer to existing ftrace
5258 instrumentation using the `TRACE_EVENT()` API.
5260 This section does not document the `TRACE_EVENT()` macro. You can
5261 read the following articles to learn more about this API:
5263 * http://lwn.net/Articles/379903/[Using the TRACE_EVENT() macro (Part 1)]
5264 * http://lwn.net/Articles/381064/[Using the TRACE_EVENT() macro (Part 2)]
5265 * http://lwn.net/Articles/383362/[Using the TRACE_EVENT() macro (Part 3)]
5267 The following procedure assumes that your ftrace tracepoints are
5268 correctly defined in their own header and that they are created in
5269 one source file using the `CREATE_TRACE_POINTS` definition.
5271 To add an LTTng layer over an existing ftrace tracepoint:
5273 . Make sure the following kernel configuration options are
5279 * `CONFIG_HIGH_RES_TIMERS`
5280 * `CONFIG_TRACEPOINTS`
5283 . Build the Linux source tree with your custom ftrace tracepoints.
5284 . Boot the resulting Linux image on your target system.
5286 Confirm that the tracepoints exist by looking for their names in the
5287 dir:{/sys/kernel/debug/tracing/events/subsys} directory, where `subsys`
5288 is your subsystem's name.
5290 . Get a copy of the latest LTTng-modules{nbsp}{revision}:
5295 $ cd $(mktemp -d) &&
5296 wget http://lttng.org/files/lttng-modules/lttng-modules-latest-2.10.tar.bz2 &&
5297 tar -xf lttng-modules-latest-2.10.tar.bz2 &&
5298 cd lttng-modules-2.10.*
5302 . In dir:{instrumentation/events/lttng-module}, relative to the root
5303 of the LTTng-modules source tree, create a header file named
5304 +__subsys__.h+ for your custom subsystem +__subsys__+ and write your
5305 LTTng-modules tracepoint definitions using the LTTng-modules
5308 Start with this template:
5312 .path:{instrumentation/events/lttng-module/my_subsys.h}
5315 #define TRACE_SYSTEM my_subsys
5317 #if !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ)
5318 #define _LTTNG_MY_SUBSYS_H
5320 #include "../../../probes/lttng-tracepoint-event.h"
5321 #include <linux/tracepoint.h>
5323 LTTNG_TRACEPOINT_EVENT(
5325 * Format is identical to TRACE_EVENT()'s version for the three
5326 * following macro parameters:
5329 TP_PROTO(int my_int, const char *my_string),
5330 TP_ARGS(my_int, my_string),
5332 /* LTTng-modules specific macros */
5334 ctf_integer(int, my_int_field, my_int)
5335 ctf_string(my_bar_field, my_bar)
5339 #endif /* !defined(_LTTNG_MY_SUBSYS_H) || defined(TRACE_HEADER_MULTI_READ) */
5341 #include "../../../probes/define_trace.h"
5345 The entries in the `TP_FIELDS()` section are the list of fields for the
5346 LTTng tracepoint. This is similar to the `TP_STRUCT__entry()` part of
5347 ftrace's `TRACE_EVENT()` macro.
5349 See <<lttng-modules-tp-fields,Tracepoint fields macros>> for a
5350 complete description of the available `ctf_*()` macros.
5352 . Create the LTTng-modules probe's kernel module C source file,
5353 +probes/lttng-probe-__subsys__.c+, where +__subsys__+ is your
5358 .path:{probes/lttng-probe-my-subsys.c}
5360 #include <linux/module.h>
5361 #include "../lttng-tracer.h"
5364 * Build-time verification of mismatch between mainline
5365 * TRACE_EVENT() arguments and the LTTng-modules adaptation
5366 * layer LTTNG_TRACEPOINT_EVENT() arguments.
5368 #include <trace/events/my_subsys.h>
5370 /* Create LTTng tracepoint probes */
5371 #define LTTNG_PACKAGE_BUILD
5372 #define CREATE_TRACE_POINTS
5373 #define TRACE_INCLUDE_PATH ../instrumentation/events/lttng-module
5375 #include "../instrumentation/events/lttng-module/my_subsys.h"
5377 MODULE_LICENSE("GPL and additional rights");
5378 MODULE_AUTHOR("Your name <your-email>");
5379 MODULE_DESCRIPTION("LTTng my_subsys probes");
5380 MODULE_VERSION(__stringify(LTTNG_MODULES_MAJOR_VERSION) "."
5381 __stringify(LTTNG_MODULES_MINOR_VERSION) "."
5382 __stringify(LTTNG_MODULES_PATCHLEVEL_VERSION)
5383 LTTNG_MODULES_EXTRAVERSION);
5387 . Edit path:{probes/KBuild} and add your new kernel module object
5388 next to the existing ones:
5392 .path:{probes/KBuild}
5396 obj-m += lttng-probe-module.o
5397 obj-m += lttng-probe-power.o
5399 obj-m += lttng-probe-my-subsys.o
5405 . Build and install the LTTng kernel modules:
5410 $ make KERNELDIR=/path/to/linux
5411 # make modules_install && depmod -a
5415 Replace `/path/to/linux` with the path to the Linux source tree where
5416 you defined and used tracepoints with ftrace's `TRACE_EVENT()` macro.
5418 Note that you can also use the
5419 <<lttng-tracepoint-event-code,`LTTNG_TRACEPOINT_EVENT_CODE()` macro>>
5420 instead of `LTTNG_TRACEPOINT_EVENT()` to use custom local variables and
5421 C code that need to be executed before the event fields are recorded.
5423 The best way to learn how to use the previous LTTng-modules macros is to
5424 inspect the existing LTTng-modules tracepoint definitions in the
5425 dir:{instrumentation/events/lttng-module} header files. Compare them
5426 with the Linux kernel mainline versions in the
5427 dir:{include/trace/events} directory of the Linux source tree.
5431 [[lttng-tracepoint-event-code]]
5432 ===== Use custom C code to access the data for tracepoint fields
5434 Although we recommended to always use the
5435 <<lttng-adaptation-layer,`LTTNG_TRACEPOINT_EVENT()`>> macro to describe
5436 the arguments and fields of an LTTng-modules tracepoint when possible,
5437 sometimes you need a more complex process to access the data that the
5438 tracer records as event record fields. In other words, you need local
5439 variables and multiple C{nbsp}statements instead of simple
5440 argument-based expressions that you pass to the
5441 <<lttng-modules-tp-fields,`ctf_*()` macros of `TP_FIELDS()`>>.
5443 You can use the `LTTNG_TRACEPOINT_EVENT_CODE()` macro instead of
5444 `LTTNG_TRACEPOINT_EVENT()` to declare custom local variables and define
5445 a block of C{nbsp}code to be executed before LTTng records the fields.
5446 The structure of this macro is:
5449 .`LTTNG_TRACEPOINT_EVENT_CODE()` macro syntax.
5451 LTTNG_TRACEPOINT_EVENT_CODE(
5453 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5454 * version for the following three macro parameters:
5457 TP_PROTO(int my_int, const char *my_string),
5458 TP_ARGS(my_int, my_string),
5460 /* Declarations of custom local variables */
5463 unsigned long b = 0;
5464 const char *name = "(undefined)";
5465 struct my_struct *my_struct;
5469 * Custom code which uses both tracepoint arguments
5470 * (in TP_ARGS()) and local variables (in TP_locvar()).
5472 * Local variables are actually members of a structure pointed
5473 * to by the special variable tp_locvar.
5477 tp_locvar->a = my_int + 17;
5478 tp_locvar->my_struct = get_my_struct_at(tp_locvar->a);
5479 tp_locvar->b = my_struct_compute_b(tp_locvar->my_struct);
5480 tp_locvar->name = my_struct_get_name(tp_locvar->my_struct);
5481 put_my_struct(tp_locvar->my_struct);
5490 * Format identical to the LTTNG_TRACEPOINT_EVENT()
5491 * version for this, except that tp_locvar members can be
5492 * used in the argument expression parameters of
5493 * the ctf_*() macros.
5496 ctf_integer(unsigned long, my_struct_b, tp_locvar->b)
5497 ctf_integer(int, my_struct_a, tp_locvar->a)
5498 ctf_string(my_string_field, my_string)
5499 ctf_string(my_struct_name, tp_locvar->name)
5504 IMPORTANT: The C code defined in `TP_code()` must not have any side
5505 effects when executed. In particular, the code must not allocate
5506 memory or get resources without deallocating this memory or putting
5507 those resources afterwards.
5510 [[instrumenting-linux-kernel-tracing]]
5511 ==== Load and unload a custom probe kernel module
5513 You must load a <<lttng-adaptation-layer,created LTTng-modules probe
5514 kernel module>> in the kernel before it can emit LTTng events.
5516 To load the default probe kernel modules and a custom probe kernel
5519 * Use the opt:lttng-sessiond(8):--extra-kmod-probes option to give extra
5520 probe modules to load when starting a root <<lttng-sessiond,session
5524 .Load the `my_subsys`, `usb`, and the default probe modules.
5528 # lttng-sessiond --extra-kmod-probes=my_subsys,usb
5533 You only need to pass the subsystem name, not the whole kernel module
5536 To load _only_ a given custom probe kernel module:
5538 * Use the opt:lttng-sessiond(8):--kmod-probes option to give the probe
5539 modules to load when starting a root session daemon:
5542 .Load only the `my_subsys` and `usb` probe modules.
5546 # lttng-sessiond --kmod-probes=my_subsys,usb
5551 To confirm that a probe module is loaded:
5558 $ lsmod | grep lttng_probe_usb
5562 To unload the loaded probe modules:
5564 * Kill the session daemon with `SIGTERM`:
5569 # pkill lttng-sessiond
5573 You can also use man:modprobe(8)'s `--remove` option if the session
5574 daemon terminates abnormally.
5577 [[controlling-tracing]]
5580 Once an application or a Linux kernel is
5581 <<instrumenting,instrumented>> for LTTng tracing,
5584 This section is divided in topics on how to use the various
5585 <<plumbing,components of LTTng>>, in particular the <<lttng-cli,cmd:lttng
5586 command-line tool>>, to _control_ the LTTng daemons and tracers.
5588 NOTE: In the following subsections, we refer to an man:lttng(1) command
5589 using its man page name. For example, instead of _Run the `create`
5590 command to..._, we use _Run the man:lttng-create(1) command to..._.
5594 === Start a session daemon
5596 In some situations, you need to run a <<lttng-sessiond,session daemon>>
5597 (man:lttng-sessiond(8)) _before_ you can use the man:lttng(1)
5600 You will see the following error when you run a command while no session
5604 Error: No session daemon is available
5607 The only command that automatically runs a session daemon is
5608 man:lttng-create(1), which you use to
5609 <<creating-destroying-tracing-sessions,create a tracing session>>. While
5610 this is most of the time the first operation that you do, sometimes it's
5611 not. Some examples are:
5613 * <<list-instrumentation-points,List the available instrumentation points>>.
5614 * <<saving-loading-tracing-session,Load a tracing session configuration>>.
5616 [[tracing-group]] Each Unix user must have its own running session
5617 daemon to trace user applications. The session daemon that the root user
5618 starts is the only one allowed to control the LTTng kernel tracer. Users
5619 that are part of the _tracing group_ can control the root session
5620 daemon. The default tracing group name is `tracing`; you can set it to
5621 something else with the opt:lttng-sessiond(8):--group option when you
5622 start the root session daemon.
5624 To start a user session daemon:
5626 * Run man:lttng-sessiond(8):
5631 $ lttng-sessiond --daemonize
5635 To start the root session daemon:
5637 * Run man:lttng-sessiond(8) as the root user:
5642 # lttng-sessiond --daemonize
5646 In both cases, remove the opt:lttng-sessiond(8):--daemonize option to
5647 start the session daemon in foreground.
5649 To stop a session daemon, use man:kill(1) on its process ID (standard
5652 Note that some Linux distributions could manage the LTTng session daemon
5653 as a service. In this case, you should use the service manager to
5654 start, restart, and stop session daemons.
5657 [[creating-destroying-tracing-sessions]]
5658 === Create and destroy a tracing session
5660 Almost all the LTTng control operations happen in the scope of
5661 a <<tracing-session,tracing session>>, which is the dialogue between the
5662 <<lttng-sessiond,session daemon>> and you.
5664 To create a tracing session with a generated name:
5666 * Use the man:lttng-create(1) command:
5675 The created tracing session's name is `auto` followed by the
5678 To create a tracing session with a specific name:
5680 * Use the optional argument of the man:lttng-create(1) command:
5685 $ lttng create my-session
5689 Replace `my-session` with the specific tracing session name.
5691 LTTng appends the creation date to the created tracing session's name.
5693 LTTng writes the traces of a tracing session in
5694 +$LTTNG_HOME/lttng-trace/__name__+ by default, where +__name__+ is the
5695 name of the tracing session. Note that the env:LTTNG_HOME environment
5696 variable defaults to `$HOME` if not set.
5698 To output LTTng traces to a non-default location:
5700 * Use the opt:lttng-create(1):--output option of the man:lttng-create(1) command:
5705 $ lttng create my-session --output=/tmp/some-directory
5709 You may create as many tracing sessions as you wish.
5711 To list all the existing tracing sessions for your Unix user:
5713 * Use the man:lttng-list(1) command:
5722 When you create a tracing session, it is set as the _current tracing
5723 session_. The following man:lttng(1) commands operate on the current
5724 tracing session when you don't specify one:
5726 [role="list-3-cols"]
5743 To change the current tracing session:
5745 * Use the man:lttng-set-session(1) command:
5750 $ lttng set-session new-session
5754 Replace `new-session` by the name of the new current tracing session.
5756 When you are done tracing in a given tracing session, you can destroy
5757 it. This operation frees the resources taken by the tracing session
5758 to destroy; it does not destroy the trace data that LTTng wrote for
5759 this tracing session.
5761 To destroy the current tracing session:
5763 * Use the man:lttng-destroy(1) command:
5773 [[list-instrumentation-points]]
5774 === List the available instrumentation points
5776 The <<lttng-sessiond,session daemon>> can query the running instrumented
5777 user applications and the Linux kernel to get a list of available
5778 instrumentation points. For the Linux kernel <<domain,tracing domain>>,
5779 they are tracepoints and system calls. For the user space tracing
5780 domain, they are tracepoints. For the other tracing domains, they are
5783 To list the available instrumentation points:
5785 * Use the man:lttng-list(1) command with the requested tracing domain's
5789 * opt:lttng-list(1):--kernel: Linux kernel tracepoints (your Unix user
5790 must be a root user, or it must be a member of the
5791 <<tracing-group,tracing group>>).
5792 * opt:lttng-list(1):--kernel with opt:lttng-list(1):--syscall: Linux
5793 kernel system calls (your Unix user must be a root user, or it must be
5794 a member of the tracing group).
5795 * opt:lttng-list(1):--userspace: user space tracepoints.
5796 * opt:lttng-list(1):--jul: `java.util.logging` loggers.
5797 * opt:lttng-list(1):--log4j: Apache log4j loggers.
5798 * opt:lttng-list(1):--python: Python loggers.
5801 .List the available user space tracepoints.
5805 $ lttng list --userspace
5809 .List the available Linux kernel system call tracepoints.
5813 $ lttng list --kernel --syscall
5818 [[enabling-disabling-events]]
5819 === Create and enable an event rule
5821 Once you <<creating-destroying-tracing-sessions,create a tracing
5822 session>>, you can create <<event,event rules>> with the
5823 man:lttng-enable-event(1) command.
5825 You specify each condition with a command-line option. The available
5826 condition options are shown in the following table.
5828 [role="growable",cols="asciidoc,asciidoc,default"]
5829 .Condition command-line options for the man:lttng-enable-event(1) command.
5831 |Option |Description |Applicable tracing domains
5837 . +--probe=__ADDR__+
5838 . +--function=__ADDR__+
5841 Instead of using the default _tracepoint_ instrumentation type, use:
5843 . A Linux system call.
5844 . A Linux https://lwn.net/Articles/132196/[KProbe] (symbol or address).
5845 . The entry and return points of a Linux function (symbol or address).
5849 |First positional argument.
5852 Tracepoint or system call name. In the case of a Linux KProbe or
5853 function, this is a custom name given to the event rule. With the
5854 JUL, log4j, and Python domains, this is a logger name.
5856 With a tracepoint, logger, or system call name, you can use the special
5857 `*` globbing character to match anything (for example, `sched_*`,
5865 . +--loglevel=__LEVEL__+
5866 . +--loglevel-only=__LEVEL__+
5869 . Match only tracepoints or log statements with a logging level at
5870 least as severe as +__LEVEL__+.
5871 . Match only tracepoints or log statements with a logging level
5872 equal to +__LEVEL__+.
5874 See man:lttng-enable-event(1) for the list of available logging level
5877 |User space, JUL, log4j, and Python.
5879 |+--exclude=__EXCLUSIONS__+
5882 When you use a `*` character at the end of the tracepoint or logger
5883 name (first positional argument), exclude the specific names in the
5884 comma-delimited list +__EXCLUSIONS__+.
5887 User space, JUL, log4j, and Python.
5889 |+--filter=__EXPR__+
5892 Match only events which satisfy the expression +__EXPR__+.
5894 See man:lttng-enable-event(1) to learn more about the syntax of a
5901 You attach an event rule to a <<channel,channel>> on creation. If you do
5902 not specify the channel with the opt:lttng-enable-event(1):--channel
5903 option, and if the event rule to create is the first in its
5904 <<domain,tracing domain>> for a given tracing session, then LTTng
5905 creates a _default channel_ for you. This default channel is reused in
5906 subsequent invocations of the man:lttng-enable-event(1) command for the
5907 same tracing domain.
5909 An event rule is always enabled at creation time.
5911 The following examples show how you can combine the previous
5912 command-line options to create simple to more complex event rules.
5914 .Create an event rule targetting a Linux kernel tracepoint (default channel).
5918 $ lttng enable-event --kernel sched_switch
5922 .Create an event rule matching four Linux kernel system calls (default channel).
5926 $ lttng enable-event --kernel --syscall open,write,read,close
5930 .Create event rules matching tracepoints with filter expressions (default channel).
5934 $ lttng enable-event --kernel sched_switch --filter='prev_comm == "bash"'
5939 $ lttng enable-event --kernel --all \
5940 --filter='$ctx.tid == 1988 || $ctx.tid == 1534'
5945 $ lttng enable-event --jul my_logger \
5946 --filter='$app.retriever:cur_msg_id > 3'
5949 IMPORTANT: Make sure to always quote the filter string when you
5950 use man:lttng(1) from a shell.
5953 .Create an event rule matching any user space tracepoint of a given tracepoint provider with a log level range (default channel).
5957 $ lttng enable-event --userspace my_app:'*' --loglevel=TRACE_INFO
5960 IMPORTANT: Make sure to always quote the wildcard character when you
5961 use man:lttng(1) from a shell.
5964 .Create an event rule matching multiple Python loggers with a wildcard and with exclusions (default channel).
5968 $ lttng enable-event --python my-app.'*' \
5969 --exclude='my-app.module,my-app.hello'
5973 .Create an event rule matching any Apache log4j logger with a specific log level (default channel).
5977 $ lttng enable-event --log4j --all --loglevel-only=LOG4J_WARN
5981 .Create an event rule attached to a specific channel matching a specific user space tracepoint provider and tracepoint.
5985 $ lttng enable-event --userspace my_app:my_tracepoint --channel=my-channel
5989 The event rules of a given channel form a whitelist: as soon as an
5990 emitted event passes one of them, LTTng can record the event. For
5991 example, an event named `my_app:my_tracepoint` emitted from a user space
5992 tracepoint with a `TRACE_ERROR` log level passes both of the following
5997 $ lttng enable-event --userspace my_app:my_tracepoint
5998 $ lttng enable-event --userspace my_app:my_tracepoint \
5999 --loglevel=TRACE_INFO
6002 The second event rule is redundant: the first one includes
6006 [[disable-event-rule]]
6007 === Disable an event rule
6009 To disable an event rule that you <<enabling-disabling-events,created>>
6010 previously, use the man:lttng-disable-event(1) command. This command
6011 disables _all_ the event rules (of a given tracing domain and channel)
6012 which match an instrumentation point. The other conditions are not
6013 supported as of LTTng{nbsp}{revision}.
6015 The LTTng tracer does not record an emitted event which passes
6016 a _disabled_ event rule.
6018 .Disable an event rule matching a Python logger (default channel).
6022 $ lttng disable-event --python my-logger
6026 .Disable an event rule matching all `java.util.logging` loggers (default channel).
6030 $ lttng disable-event --jul '*'
6034 .Disable _all_ the event rules of the default channel.
6036 The opt:lttng-disable-event(1):--all-events option is not, like the
6037 opt:lttng-enable-event(1):--all option of man:lttng-enable-event(1), the
6038 equivalent of the event name `*` (wildcard): it disables _all_ the event
6039 rules of a given channel.
6043 $ lttng disable-event --jul --all-events
6047 NOTE: You cannot delete an event rule once you create it.
6051 === Get the status of a tracing session
6053 To get the status of the current tracing session, that is, its
6054 parameters, its channels, event rules, and their attributes:
6056 * Use the man:lttng-status(1) command:
6066 To get the status of any tracing session:
6068 * Use the man:lttng-list(1) command with the tracing session's name:
6073 $ lttng list my-session
6077 Replace `my-session` with the desired tracing session's name.
6080 [[basic-tracing-session-control]]
6081 === Start and stop a tracing session
6083 Once you <<creating-destroying-tracing-sessions,create a tracing
6085 <<enabling-disabling-events,create one or more event rules>>,
6086 you can start and stop the tracers for this tracing session.
6088 To start tracing in the current tracing session:
6090 * Use the man:lttng-start(1) command:
6099 LTTng is very flexible: you can launch user applications before
6100 or after the you start the tracers. The tracers only record the events
6101 if they pass enabled event rules and if they occur while the tracers are
6104 To stop tracing in the current tracing session:
6106 * Use the man:lttng-stop(1) command:
6115 If there were <<channel-overwrite-mode-vs-discard-mode,lost event
6116 records>> or lost sub-buffers since the last time you ran
6117 man:lttng-start(1), warnings are printed when you run the
6118 man:lttng-stop(1) command.
6120 IMPORTANT: You need to stop tracing to make LTTng flush the remaining
6121 trace data and make the trace readable.
6124 [[enabling-disabling-channels]]
6125 === Create a channel
6127 Once you create a tracing session, you can create a <<channel,channel>>
6128 with the man:lttng-enable-channel(1) command.
6130 Note that LTTng automatically creates a default channel when, for a
6131 given <<domain,tracing domain>>, no channels exist and you
6132 <<enabling-disabling-events,create>> the first event rule. This default
6133 channel is named `channel0` and its attributes are set to reasonable
6134 values. Therefore, you only need to create a channel when you need
6135 non-default attributes.
6137 You specify each non-default channel attribute with a command-line
6138 option when you use the man:lttng-enable-channel(1) command. The
6139 available command-line options are:
6141 [role="growable",cols="asciidoc,asciidoc"]
6142 .Command-line options for the man:lttng-enable-channel(1) command.
6144 |Option |Description
6150 <<channel-overwrite-mode-vs-discard-mode,event loss mode>> instead of
6151 the default _discard_ mode.
6153 |`--buffers-pid` (user space tracing domain only)
6156 Use the per-process <<channel-buffering-schemes,buffering scheme>>
6157 instead of the default per-user buffering scheme.
6159 |+--subbuf-size=__SIZE__+
6162 Allocate sub-buffers of +__SIZE__+ bytes (power of two), for each CPU,
6163 either for each Unix user (default), or for each instrumented process.
6165 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6167 |+--num-subbuf=__COUNT__+
6170 Allocate +__COUNT__+ sub-buffers (power of two), for each CPU, either
6171 for each Unix user (default), or for each instrumented process.
6173 See <<channel-subbuf-size-vs-subbuf-count,Sub-buffer count and size>>.
6175 |+--tracefile-size=__SIZE__+
6178 Set the maximum size of each trace file that this channel writes within
6179 a stream to +__SIZE__+ bytes instead of no maximum.
6181 See <<tracefile-rotation,Trace file count and size>>.
6183 |+--tracefile-count=__COUNT__+
6186 Limit the number of trace files that this channel creates to
6187 +__COUNT__+ channels instead of no limit.
6189 See <<tracefile-rotation,Trace file count and size>>.
6191 |+--switch-timer=__PERIODUS__+
6194 Set the <<channel-switch-timer,switch timer period>>
6195 to +__PERIODUS__+{nbsp}µs.
6197 |+--read-timer=__PERIODUS__+
6200 Set the <<channel-read-timer,read timer period>>
6201 to +__PERIODUS__+{nbsp}µs.
6203 |[[opt-blocking-timeout]]+--blocking-timeout=__TIMEOUTUS__+
6206 Set the timeout of user space applications which load LTTng-UST
6207 in blocking mode to +__TIMEOUTUS__+:
6210 Never block (non-blocking mode).
6213 Block forever until space is available in a sub-buffer to record
6216 __n__, a positive value::
6217 Wait for at most __n__ µs when trying to write into a sub-buffer.
6219 Note that, for this option to have any effect on an instrumented
6220 user space application, you need to run the application with a set
6221 env:LTTNG_UST_ALLOW_BLOCKING environment variable.
6223 |+--output=__TYPE__+ (Linux kernel tracing domain only)
6226 Set the channel's output type to +__TYPE__+, either `mmap` or `splice`.
6230 You can only create a channel in the Linux kernel and user space
6231 <<domain,tracing domains>>: other tracing domains have their own channel
6232 created on the fly when <<enabling-disabling-events,creating event
6237 Because of a current LTTng limitation, you must create all channels
6238 _before_ you <<basic-tracing-session-control,start tracing>> in a given
6239 tracing session, that is, before the first time you run
6242 Since LTTng automatically creates a default channel when you use the
6243 man:lttng-enable-event(1) command with a specific tracing domain, you
6244 cannot, for example, create a Linux kernel event rule, start tracing,
6245 and then create a user space event rule, because no user space channel
6246 exists yet and it's too late to create one.
6248 For this reason, make sure to configure your channels properly
6249 before starting the tracers for the first time!
6252 The following examples show how you can combine the previous
6253 command-line options to create simple to more complex channels.
6255 .Create a Linux kernel channel with default attributes.
6259 $ lttng enable-channel --kernel my-channel
6263 .Create a user space channel with 4 sub-buffers or 1{nbsp}MiB each, per CPU, per instrumented process.
6267 $ lttng enable-channel --userspace --num-subbuf=4 --subbuf-size=1M \
6268 --buffers-pid my-channel
6272 .[[blocking-timeout-example]]Create a default user space channel with an infinite blocking timeout.
6274 <<creating-destroying-tracing-sessions,Create a tracing-session>>,
6275 create the channel, <<enabling-disabling-events,create an event rule>>,
6276 and <<basic-tracing-session-control,start tracing>>:
6281 $ lttng enable-channel --userspace --blocking-timeout=inf blocking-channel
6282 $ lttng enable-event --userspace --channel=blocking-channel --all
6286 Run an application instrumented with LTTng-UST and allow it to block:
6290 $ LTTNG_UST_ALLOW_BLOCKING=1 my-app
6294 .Create a Linux kernel channel which rotates 8 trace files of 4{nbsp}MiB each for each stream
6298 $ lttng enable-channel --kernel --tracefile-count=8 \
6299 --tracefile-size=4194304 my-channel
6303 .Create a user space channel in overwrite (or _flight recorder_) mode.
6307 $ lttng enable-channel --userspace --overwrite my-channel
6311 You can <<enabling-disabling-events,create>> the same event rule in
6312 two different channels:
6316 $ lttng enable-event --userspace --channel=my-channel app:tp
6317 $ lttng enable-event --userspace --channel=other-channel app:tp
6320 If both channels are enabled, when a tracepoint named `app:tp` is
6321 reached, LTTng records two events, one for each channel.
6325 === Disable a channel
6327 To disable a specific channel that you <<enabling-disabling-channels,created>>
6328 previously, use the man:lttng-disable-channel(1) command.
6330 .Disable a specific Linux kernel channel.
6334 $ lttng disable-channel --kernel my-channel
6338 The state of a channel precedes the individual states of event rules
6339 attached to it: event rules which belong to a disabled channel, even if
6340 they are enabled, are also considered disabled.
6344 === Add context fields to a channel
6346 Event record fields in trace files provide important information about
6347 events that occured previously, but sometimes some external context may
6348 help you solve a problem faster. Examples of context fields are:
6350 * The **process ID**, **thread ID**, **process name**, and
6351 **process priority** of the thread in which the event occurs.
6352 * The **hostname** of the system on which the event occurs.
6353 * The current values of many possible **performance counters** using
6355 ** CPU cycles, stalled cycles, idle cycles, and the other cycle types.
6357 ** Branch instructions, misses, and loads.
6359 * Any context defined at the application level (supported for the
6360 JUL and log4j <<domain,tracing domains>>).
6362 To get the full list of available context fields, see
6363 `lttng add-context --list`. Some context fields are reserved for a
6364 specific <<domain,tracing domain>> (Linux kernel or user space).
6366 You add context fields to <<channel,channels>>. All the events
6367 that a channel with added context fields records contain those fields.
6369 To add context fields to one or all the channels of a given tracing
6372 * Use the man:lttng-add-context(1) command.
6374 .Add context fields to all the channels of the current tracing session.
6376 The following command line adds the virtual process identifier and
6377 the per-thread CPU cycles count fields to all the user space channels
6378 of the current tracing session.
6382 $ lttng add-context --userspace --type=vpid --type=perf:thread:cpu-cycles
6386 .Add performance counter context fields by raw ID
6388 See man:lttng-add-context(1) for the exact format of the context field
6389 type, which is partly compatible with the format used in
6394 $ lttng add-context --userspace --type=perf:thread:raw:r0110:test
6395 $ lttng add-context --kernel --type=perf:cpu:raw:r0013c:x86unhalted
6399 .Add a context field to a specific channel.
6401 The following command line adds the thread identifier context field
6402 to the Linux kernel channel named `my-channel` in the current
6407 $ lttng add-context --kernel --channel=my-channel --type=tid
6411 .Add an application-specific context field to a specific channel.
6413 The following command line adds the `cur_msg_id` context field of the
6414 `retriever` context retriever for all the instrumented
6415 <<java-application,Java applications>> recording <<event,event records>>
6416 in the channel named `my-channel`:
6420 $ lttng add-context --kernel --channel=my-channel \
6421 --type='$app:retriever:cur_msg_id'
6424 IMPORTANT: Make sure to always quote the `$` character when you
6425 use man:lttng-add-context(1) from a shell.
6428 NOTE: You cannot remove context fields from a channel once you add it.
6433 === Track process IDs
6435 It's often useful to allow only specific process IDs (PIDs) to emit
6436 events. For example, you may wish to record all the system calls made by
6437 a given process (Ă la http://linux.die.net/man/1/strace[strace]).
6439 The man:lttng-track(1) and man:lttng-untrack(1) commands serve this
6440 purpose. Both commands operate on a whitelist of process IDs. You _add_
6441 entries to this whitelist with the man:lttng-track(1) command and remove
6442 entries with the man:lttng-untrack(1) command. Any process which has one
6443 of the PIDs in the whitelist is allowed to emit LTTng events which pass
6444 an enabled <<event,event rule>>.
6446 NOTE: The PID tracker tracks the _numeric process IDs_. Should a
6447 process with a given tracked ID exit and another process be given this
6448 ID, then the latter would also be allowed to emit events.
6450 .Track and untrack process IDs.
6452 For the sake of the following example, assume the target system has 16
6456 <<creating-destroying-tracing-sessions,create a tracing session>>,
6457 the whitelist contains all the possible PIDs:
6460 .All PIDs are tracked.
6461 image::track-all.png[]
6463 When the whitelist is full and you use the man:lttng-track(1) command to
6464 specify some PIDs to track, LTTng first clears the whitelist, then it
6465 tracks the specific PIDs. After:
6469 $ lttng track --pid=3,4,7,10,13
6475 .PIDs 3, 4, 7, 10, and 13 are tracked.
6476 image::track-3-4-7-10-13.png[]
6478 You can add more PIDs to the whitelist afterwards:
6482 $ lttng track --pid=1,15,16
6488 .PIDs 1, 15, and 16 are added to the whitelist.
6489 image::track-1-3-4-7-10-13-15-16.png[]
6491 The man:lttng-untrack(1) command removes entries from the PID tracker's
6492 whitelist. Given the previous example, the following command:
6496 $ lttng untrack --pid=3,7,10,13
6499 leads to this whitelist:
6502 .PIDs 3, 7, 10, and 13 are removed from the whitelist.
6503 image::track-1-4-15-16.png[]
6505 LTTng can track all possible PIDs again using the
6506 opt:lttng-track(1):--all option:
6510 $ lttng track --pid --all
6513 The result is, again:
6516 .All PIDs are tracked.
6517 image::track-all.png[]
6520 .Track only specific PIDs
6522 A very typical use case with PID tracking is to start with an empty
6523 whitelist, then <<basic-tracing-session-control,start the tracers>>, and
6524 then add PIDs manually while tracers are active. You can accomplish this
6525 by using the opt:lttng-untrack(1):--all option of the
6526 man:lttng-untrack(1) command to clear the whitelist after you
6527 <<creating-destroying-tracing-sessions,create a tracing session>>:
6531 $ lttng untrack --pid --all
6537 .No PIDs are tracked.
6538 image::untrack-all.png[]
6540 If you trace with this whitelist configuration, the tracer records no
6541 events for this <<domain,tracing domain>> because no processes are
6542 tracked. You can use the man:lttng-track(1) command as usual to track
6543 specific PIDs, for example:
6547 $ lttng track --pid=6,11
6553 .PIDs 6 and 11 are tracked.
6554 image::track-6-11.png[]
6559 [[saving-loading-tracing-session]]
6560 === Save and load tracing session configurations
6562 Configuring a <<tracing-session,tracing session>> can be long. Some of
6563 the tasks involved are:
6565 * <<enabling-disabling-channels,Create channels>> with
6566 specific attributes.
6567 * <<adding-context,Add context fields>> to specific channels.
6568 * <<enabling-disabling-events,Create event rules>> with specific log
6569 level and filter conditions.
6571 If you use LTTng to solve real world problems, chances are you have to
6572 record events using the same tracing session setup over and over,
6573 modifying a few variables each time in your instrumented program
6574 or environment. To avoid constant tracing session reconfiguration,
6575 the man:lttng(1) command-line tool can save and load tracing session
6576 configurations to/from XML files.
6578 To save a given tracing session configuration:
6580 * Use the man:lttng-save(1) command:
6585 $ lttng save my-session
6589 Replace `my-session` with the name of the tracing session to save.
6591 LTTng saves tracing session configurations to
6592 dir:{$LTTNG_HOME/.lttng/sessions} by default. Note that the
6593 env:LTTNG_HOME environment variable defaults to `$HOME` if not set. Use
6594 the opt:lttng-save(1):--output-path option to change this destination
6597 LTTng saves all configuration parameters, for example:
6599 * The tracing session name.
6600 * The trace data output path.
6601 * The channels with their state and all their attributes.
6602 * The context fields you added to channels.
6603 * The event rules with their state, log level and filter conditions.
6605 To load a tracing session:
6607 * Use the man:lttng-load(1) command:
6612 $ lttng load my-session
6616 Replace `my-session` with the name of the tracing session to load.
6618 When LTTng loads a configuration, it restores your saved tracing session
6619 as if you just configured it manually.
6621 See man:lttng(1) for the complete list of command-line options. You
6622 can also save and load all many sessions at a time, and decide in which
6623 directory to output the XML files.
6626 [[sending-trace-data-over-the-network]]
6627 === Send trace data over the network
6629 LTTng can send the recorded trace data to a remote system over the
6630 network instead of writing it to the local file system.
6632 To send the trace data over the network:
6634 . On the _remote_ system (which can also be the target system),
6635 start an LTTng <<lttng-relayd,relay daemon>> (man:lttng-relayd(8)):
6644 . On the _target_ system, create a tracing session configured to
6645 send trace data over the network:
6650 $ lttng create my-session --set-url=net://remote-system
6654 Replace `remote-system` by the host name or IP address of the
6655 remote system. See man:lttng-create(1) for the exact URL format.
6657 . On the target system, use the man:lttng(1) command-line tool as usual.
6658 When tracing is active, the target's consumer daemon sends sub-buffers
6659 to the relay daemon running on the remote system instead of flushing
6660 them to the local file system. The relay daemon writes the received
6661 packets to the local file system.
6663 The relay daemon writes trace files to
6664 +$LTTNG_HOME/lttng-traces/__hostname__/__session__+ by default, where
6665 +__hostname__+ is the host name of the target system and +__session__+
6666 is the tracing session name. Note that the env:LTTNG_HOME environment
6667 variable defaults to `$HOME` if not set. Use the
6668 opt:lttng-relayd(8):--output option of man:lttng-relayd(8) to write
6669 trace files to another base directory.
6674 === View events as LTTng emits them (noch:{LTTng} live)
6676 LTTng live is a network protocol implemented by the <<lttng-relayd,relay
6677 daemon>> (man:lttng-relayd(8)) to allow compatible trace viewers to
6678 display events as LTTng emits them on the target system while tracing is
6681 The relay daemon creates a _tee_: it forwards the trace data to both
6682 the local file system and to connected live viewers:
6685 .The relay daemon creates a _tee_, forwarding the trace data to both trace files and a connected live viewer.
6690 . On the _target system_, create a <<tracing-session,tracing session>>
6696 $ lttng create my-session --live
6700 This spawns a local relay daemon.
6702 . Start the live viewer and configure it to connect to the relay
6703 daemon. For example, with http://diamon.org/babeltrace[Babeltrace]:
6708 $ babeltrace --input-format=lttng-live \
6709 net://localhost/host/hostname/my-session
6716 * `hostname` with the host name of the target system.
6717 * `my-session` with the name of the tracing session to view.
6720 . Configure the tracing session as usual with the man:lttng(1)
6721 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6723 You can list the available live tracing sessions with Babeltrace:
6727 $ babeltrace --input-format=lttng-live net://localhost
6730 You can start the relay daemon on another system. In this case, you need
6731 to specify the relay daemon's URL when you create the tracing session
6732 with the opt:lttng-create(1):--set-url option. You also need to replace
6733 `localhost` in the procedure above with the host name of the system on
6734 which the relay daemon is running.
6736 See man:lttng-create(1) and man:lttng-relayd(8) for the complete list of
6737 command-line options.
6741 [[taking-a-snapshot]]
6742 === Take a snapshot of the current sub-buffers of a tracing session
6744 The normal behavior of LTTng is to append full sub-buffers to growing
6745 trace data files. This is ideal to keep a full history of the events
6746 that occurred on the target system, but it can
6747 represent too much data in some situations. For example, you may wish
6748 to trace your application continuously until some critical situation
6749 happens, in which case you only need the latest few recorded
6750 events to perform the desired analysis, not multi-gigabyte trace files.
6752 With the man:lttng-snapshot(1) command, you can take a snapshot of the
6753 current sub-buffers of a given <<tracing-session,tracing session>>.
6754 LTTng can write the snapshot to the local file system or send it over
6759 . Create a tracing session in _snapshot mode_:
6764 $ lttng create my-session --snapshot
6768 The <<channel-overwrite-mode-vs-discard-mode,event loss mode>> of
6769 <<channel,channels>> created in this mode is automatically set to
6770 _overwrite_ (flight recorder mode).
6772 . Configure the tracing session as usual with the man:lttng(1)
6773 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6775 . **Optional**: When you need to take a snapshot,
6776 <<basic-tracing-session-control,stop tracing>>.
6778 You can take a snapshot when the tracers are active, but if you stop
6779 them first, you are sure that the data in the sub-buffers does not
6780 change before you actually take the snapshot.
6787 $ lttng snapshot record --name=my-first-snapshot
6791 LTTng writes the current sub-buffers of all the current tracing
6792 session's channels to trace files on the local file system. Those trace
6793 files have `my-first-snapshot` in their name.
6795 There is no difference between the format of a normal trace file and the
6796 format of a snapshot: viewers of LTTng traces also support LTTng
6799 By default, LTTng writes snapshot files to the path shown by
6800 `lttng snapshot list-output`. You can change this path or decide to send
6801 snapshots over the network using either:
6803 . An output path or URL that you specify when you create the
6805 . An snapshot output path or URL that you add using
6806 `lttng snapshot add-output`
6807 . An output path or URL that you provide directly to the
6808 `lttng snapshot record` command.
6810 Method 3 overrides method 2, which overrides method 1. When you
6811 specify a URL, a relay daemon must listen on a remote system (see
6812 <<sending-trace-data-over-the-network,Send trace data over the network>>).
6817 === Use the machine interface
6819 With any command of the man:lttng(1) command-line tool, you can set the
6820 opt:lttng(1):--mi option to `xml` (before the command name) to get an
6821 XML machine interface output, for example:
6825 $ lttng --mi=xml enable-event --kernel --syscall open
6828 A schema definition (XSD) is
6829 https://github.com/lttng/lttng-tools/blob/stable-2.10/src/common/mi-lttng-3.0.xsd[available]
6830 to ease the integration with external tools as much as possible.
6834 [[metadata-regenerate]]
6835 === Regenerate the metadata of an LTTng trace
6837 An LTTng trace, which is a http://diamon.org/ctf[CTF] trace, has both
6838 data stream files and a metadata file. This metadata file contains,
6839 amongst other things, information about the offset of the clock sources
6840 used to timestamp <<event,event records>> when tracing.
6842 If, once a <<tracing-session,tracing session>> is
6843 <<basic-tracing-session-control,started>>, a major
6844 https://en.wikipedia.org/wiki/Network_Time_Protocol[NTP] correction
6845 happens, the trace's clock offset also needs to be updated. You
6846 can use the `metadata` item of the man:lttng-regenerate(1) command
6849 The main use case of this command is to allow a system to boot with
6850 an incorrect wall time and trace it with LTTng before its wall time
6851 is corrected. Once the system is known to be in a state where its
6852 wall time is correct, it can run `lttng regenerate metadata`.
6854 To regenerate the metadata of an LTTng trace:
6856 * Use the `metadata` item of the man:lttng-regenerate(1) command:
6861 $ lttng regenerate metadata
6867 `lttng regenerate metadata` has the following limitations:
6869 * Tracing session <<creating-destroying-tracing-sessions,created>>
6871 * User space <<channel,channels>>, if any, are using
6872 <<channel-buffering-schemes,per-user buffering>>.
6877 [[regenerate-statedump]]
6878 === Regenerate the state dump of a tracing session
6880 The LTTng kernel and user space tracers generate state dump
6881 <<event,event records>> when the application starts or when you
6882 <<basic-tracing-session-control,start a tracing session>>. An analysis
6883 can use the state dump event records to set an initial state before it
6884 builds the rest of the state from the following event records.
6885 http://tracecompass.org/[Trace Compass] is a notable example of an
6886 application which uses the state dump of an LTTng trace.
6888 When you <<taking-a-snapshot,take a snapshot>>, it's possible that the
6889 state dump event records are not included in the snapshot because they
6890 were recorded to a sub-buffer that has been consumed or overwritten
6893 You can use the `lttng regenerate statedump` command to emit the state
6894 dump event records again.
6896 To regenerate the state dump of the current tracing session, provided
6897 create it in snapshot mode, before you take a snapshot:
6899 . Use the `statedump` item of the man:lttng-regenerate(1) command:
6904 $ lttng regenerate statedump
6908 . <<basic-tracing-session-control,Stop the tracing session>>:
6917 . <<taking-a-snapshot,Take a snapshot>>:
6922 $ lttng snapshot record --name=my-snapshot
6926 Depending on the event throughput, you should run steps 1 and 2
6927 as closely as possible.
6929 NOTE: To record the state dump events, you need to
6930 <<enabling-disabling-events,create event rules>> which enable them.
6931 LTTng-UST state dump tracepoints start with `lttng_ust_statedump:`.
6932 LTTng-modules state dump tracepoints start with `lttng_statedump_`.
6936 [[persistent-memory-file-systems]]
6937 === Record trace data on persistent memory file systems
6939 https://en.wikipedia.org/wiki/Non-volatile_random-access_memory[Non-volatile random-access memory]
6940 (NVRAM) is random-access memory that retains its information when power
6941 is turned off (non-volatile). Systems with such memory can store data
6942 structures in RAM and retrieve them after a reboot, without flushing
6943 to typical _storage_.
6945 Linux supports NVRAM file systems thanks to either
6946 http://pramfs.sourceforge.net/[PRAMFS] or
6947 https://www.kernel.org/doc/Documentation/filesystems/dax.txt[DAX]{nbsp}+{nbsp}http://lkml.iu.edu/hypermail/linux/kernel/1504.1/03463.html[pmem]
6948 (requires Linux 4.1+).
6950 This section does not describe how to operate such file systems;
6951 we assume that you have a working persistent memory file system.
6953 When you create a <<tracing-session,tracing session>>, you can specify
6954 the path of the shared memory holding the sub-buffers. If you specify a
6955 location on an NVRAM file system, then you can retrieve the latest
6956 recorded trace data when the system reboots after a crash.
6958 To record trace data on a persistent memory file system and retrieve the
6959 trace data after a system crash:
6961 . Create a tracing session with a sub-buffer shared memory path located
6962 on an NVRAM file system:
6967 $ lttng create my-session --shm-path=/path/to/shm
6971 . Configure the tracing session as usual with the man:lttng(1)
6972 command-line tool, and <<basic-tracing-session-control,start tracing>>.
6974 . After a system crash, use the man:lttng-crash(1) command-line tool to
6975 view the trace data recorded on the NVRAM file system:
6980 $ lttng-crash /path/to/shm
6984 The binary layout of the ring buffer files is not exactly the same as
6985 the trace files layout. This is why you need to use man:lttng-crash(1)
6986 instead of your preferred trace viewer directly.
6988 To convert the ring buffer files to LTTng trace files:
6990 * Use the opt:lttng-crash(1):--extract option of man:lttng-crash(1):
6995 $ lttng-crash --extract=/path/to/trace /path/to/shm
7001 [[notif-trigger-api]]
7002 === Get notified when a channel's buffer usage is too high or too low
7004 With LTTng's $$C/C++$$ notification and trigger API, your user
7005 application can get notified when the buffer usage of one or more
7006 <<channel,channels>> becomes too low or too high. You can use this API
7007 and enable or disable <<event,event rules>> during tracing to avoid
7008 <<channel-overwrite-mode-vs-discard-mode,discarded event records>>.
7010 .Have a user application get notified when an LTTng channel's buffer usage is too high.
7012 In this example, we create and build an application which gets notified
7013 when the buffer usage of a specific LTTng channel is higher than
7014 75{nbsp}%. We only print that it is the case in the example, but we
7015 could as well use the API of <<liblttng-ctl-lttng,`liblttng-ctl`>> to
7016 disable event rules when this happens.
7018 . Create the application's C source file:
7026 #include <lttng/domain.h>
7027 #include <lttng/action/action.h>
7028 #include <lttng/action/notify.h>
7029 #include <lttng/condition/condition.h>
7030 #include <lttng/condition/buffer-usage.h>
7031 #include <lttng/condition/evaluation.h>
7032 #include <lttng/notification/channel.h>
7033 #include <lttng/notification/notification.h>
7034 #include <lttng/trigger/trigger.h>
7035 #include <lttng/endpoint.h>
7037 int main(int argc, char *argv[])
7039 int exit_status = 0;
7040 struct lttng_notification_channel *notification_channel;
7041 struct lttng_condition *condition;
7042 struct lttng_action *action;
7043 struct lttng_trigger *trigger;
7044 const char *tracing_session_name;
7045 const char *channel_name;
7048 tracing_session_name = argv[1];
7049 channel_name = argv[2];
7052 * Create a notification channel. A notification channel
7053 * connects the user application to the LTTng session daemon.
7054 * This notification channel can be used to listen to various
7055 * types of notifications.
7057 notification_channel = lttng_notification_channel_create(
7058 lttng_session_daemon_notification_endpoint);
7061 * Create a "high buffer usage" condition. In this case, the
7062 * condition is reached when the buffer usage is greater than or
7063 * equal to 75 %. We create the condition for a specific tracing
7064 * session name, channel name, and for the user space tracing
7067 * The "low buffer usage" condition type also exists.
7069 condition = lttng_condition_buffer_usage_high_create();
7070 lttng_condition_buffer_usage_set_threshold_ratio(condition, .75);
7071 lttng_condition_buffer_usage_set_session_name(
7072 condition, tracing_session_name);
7073 lttng_condition_buffer_usage_set_channel_name(condition,
7075 lttng_condition_buffer_usage_set_domain_type(condition,
7079 * Create an action (get a notification) to take when the
7080 * condition created above is reached.
7082 action = lttng_action_notify_create();
7085 * Create a trigger. A trigger associates a condition to an
7086 * action: the action is executed when the condition is reached.
7088 trigger = lttng_trigger_create(condition, action);
7090 /* Register the trigger to LTTng. */
7091 lttng_register_trigger(trigger);
7094 * Now that we have registered a trigger, a notification will be
7095 * emitted everytime its condition is met. To receive this
7096 * notification, we must subscribe to notifications that match
7097 * the same condition.
7099 lttng_notification_channel_subscribe(notification_channel,
7103 * Notification loop. You can put this in a dedicated thread to
7104 * avoid blocking the main thread.
7107 struct lttng_notification *notification;
7108 enum lttng_notification_channel_status status;
7109 const struct lttng_evaluation *notification_evaluation;
7110 const struct lttng_condition *notification_condition;
7111 double buffer_usage;
7113 /* Receive the next notification. */
7114 status = lttng_notification_channel_get_next_notification(
7115 notification_channel, ¬ification);
7118 case LTTNG_NOTIFICATION_CHANNEL_STATUS_OK:
7120 case LTTNG_NOTIFICATION_CHANNEL_STATUS_NOTIFICATIONS_DROPPED:
7122 * The session daemon can drop notifications if
7123 * a monitoring application is not consuming the
7124 * notifications fast enough.
7127 case LTTNG_NOTIFICATION_CHANNEL_STATUS_CLOSED:
7129 * The notification channel has been closed by the
7130 * session daemon. This is typically caused by a session
7131 * daemon shutting down.
7135 /* Unhandled conditions or errors. */
7141 * A notification provides, amongst other things:
7143 * * The condition that caused this notification to be
7145 * * The condition evaluation, which provides more
7146 * specific information on the evaluation of the
7149 * The condition evaluation provides the buffer usage
7150 * value at the moment the condition was reached.
7152 notification_condition = lttng_notification_get_condition(
7154 notification_evaluation = lttng_notification_get_evaluation(
7157 /* We're subscribed to only one condition. */
7158 assert(lttng_condition_get_type(notification_condition) ==
7159 LTTNG_CONDITION_TYPE_BUFFER_USAGE_HIGH);
7162 * Get the exact sampled buffer usage from the
7163 * condition evaluation.
7165 lttng_evaluation_buffer_usage_get_usage_ratio(
7166 notification_evaluation, &buffer_usage);
7169 * At this point, instead of printing a message, we
7170 * could do something to reduce the channel's buffer
7171 * usage, like disable specific events.
7173 printf("Buffer usage is %f %% in tracing session \"%s\", "
7174 "user space channel \"%s\".\n", buffer_usage * 100,
7175 tracing_session_name, channel_name);
7176 lttng_notification_destroy(notification);
7180 lttng_action_destroy(action);
7181 lttng_condition_destroy(condition);
7182 lttng_trigger_destroy(trigger);
7183 lttng_notification_channel_destroy(notification_channel);
7189 . Build the `notif-app` application, linking it to `liblttng-ctl`:
7194 $ gcc -o notif-app notif-app.c -llttng-ctl
7198 . <<creating-destroying-tracing-sessions,Create a tracing session>>,
7199 <<enabling-disabling-events,create an event rule>> matching all the
7200 user space tracepoints, and
7201 <<basic-tracing-session-control,start tracing>>:
7206 $ lttng create my-session
7207 $ lttng enable-event --userspace --all
7212 If you create the channel manually with the man:lttng-enable-channel(1)
7213 command, you can control how frequently are the current values of the
7214 channel's properties sampled to evaluate user conditions with the
7215 opt:lttng-enable-channel(1):--monitor-timer option.
7217 . Run the `notif-app` application. This program accepts the
7218 <<tracing-session,tracing session>> name and the user space channel
7219 name as its two first arguments. The channel which LTTng automatically
7220 creates with the man:lttng-enable-event(1) command above is named
7226 $ ./notif-app my-session channel0
7230 . In another terminal, run an application with a very high event
7231 throughput so that the 75{nbsp}% buffer usage condition is reached.
7233 In the first terminal, the application should print lines like this:
7236 Buffer usage is 81.45197 % in tracing session "my-session", user space
7240 If you don't see anything, try modifying the condition in
7241 path:{notif-app.c} to a lower value (0.1, for example), rebuilding it
7242 (step 2) and running it again (step 4).
7249 [[lttng-modules-ref]]
7250 === noch:{LTTng-modules}
7254 [[lttng-tracepoint-enum]]
7255 ==== `LTTNG_TRACEPOINT_ENUM()` usage
7257 Use the `LTTNG_TRACEPOINT_ENUM()` macro to define an enumeration:
7261 LTTNG_TRACEPOINT_ENUM(name, TP_ENUM_VALUES(entries))
7266 * `name` with the name of the enumeration (C identifier, unique
7267 amongst all the defined enumerations).
7268 * `entries` with a list of enumeration entries.
7270 The available enumeration entry macros are:
7272 +ctf_enum_value(__name__, __value__)+::
7273 Entry named +__name__+ mapped to the integral value +__value__+.
7275 +ctf_enum_range(__name__, __begin__, __end__)+::
7276 Entry named +__name__+ mapped to the range of integral values between
7277 +__begin__+ (included) and +__end__+ (included).
7279 +ctf_enum_auto(__name__)+::
7280 Entry named +__name__+ mapped to the integral value following the
7281 last mapping's value.
7283 The last value of a `ctf_enum_value()` entry is its +__value__+
7286 The last value of a `ctf_enum_range()` entry is its +__end__+ parameter.
7288 If `ctf_enum_auto()` is the first entry in the list, its integral
7291 Use the `ctf_enum()` <<lttng-modules-tp-fields,field definition macro>>
7292 to use a defined enumeration as a tracepoint field.
7294 .Define an enumeration with `LTTNG_TRACEPOINT_ENUM()`.
7298 LTTNG_TRACEPOINT_ENUM(
7301 ctf_enum_auto("AUTO: EXPECT 0")
7302 ctf_enum_value("VALUE: 23", 23)
7303 ctf_enum_value("VALUE: 27", 27)
7304 ctf_enum_auto("AUTO: EXPECT 28")
7305 ctf_enum_range("RANGE: 101 TO 303", 101, 303)
7306 ctf_enum_auto("AUTO: EXPECT 304")
7314 [[lttng-modules-tp-fields]]
7315 ==== Tracepoint fields macros (for `TP_FIELDS()`)
7317 [[tp-fast-assign]][[tp-struct-entry]]The available macros to define
7318 tracepoint fields, which must be listed within `TP_FIELDS()` in
7319 `LTTNG_TRACEPOINT_EVENT()`, are:
7321 [role="func-desc growable",cols="asciidoc,asciidoc"]
7322 .Available macros to define LTTng-modules tracepoint fields
7324 |Macro |Description and parameters
7327 +ctf_integer(__t__, __n__, __e__)+
7329 +ctf_integer_nowrite(__t__, __n__, __e__)+
7331 +ctf_user_integer(__t__, __n__, __e__)+
7333 +ctf_user_integer_nowrite(__t__, __n__, __e__)+
7335 Standard integer, displayed in base 10.
7338 Integer C type (`int`, `long`, `size_t`, ...).
7344 Argument expression.
7347 +ctf_integer_hex(__t__, __n__, __e__)+
7349 +ctf_user_integer_hex(__t__, __n__, __e__)+
7351 Standard integer, displayed in base 16.
7360 Argument expression.
7362 |+ctf_integer_oct(__t__, __n__, __e__)+
7364 Standard integer, displayed in base 8.
7373 Argument expression.
7376 +ctf_integer_network(__t__, __n__, __e__)+
7378 +ctf_user_integer_network(__t__, __n__, __e__)+
7380 Integer in network byte order (big-endian), displayed in base 10.
7389 Argument expression.
7392 +ctf_integer_network_hex(__t__, __n__, __e__)+
7394 +ctf_user_integer_network_hex(__t__, __n__, __e__)+
7396 Integer in network byte order, displayed in base 16.
7405 Argument expression.
7408 +ctf_enum(__N__, __t__, __n__, __e__)+
7410 +ctf_enum_nowrite(__N__, __t__, __n__, __e__)+
7412 +ctf_user_enum(__N__, __t__, __n__, __e__)+
7414 +ctf_user_enum_nowrite(__N__, __t__, __n__, __e__)+
7419 Name of a <<lttng-tracepoint-enum,previously defined enumeration>>.
7422 Integer C type (`int`, `long`, `size_t`, ...).
7428 Argument expression.
7431 +ctf_string(__n__, __e__)+
7433 +ctf_string_nowrite(__n__, __e__)+
7435 +ctf_user_string(__n__, __e__)+
7437 +ctf_user_string_nowrite(__n__, __e__)+
7439 Null-terminated string; undefined behavior if +__e__+ is `NULL`.
7445 Argument expression.
7448 +ctf_array(__t__, __n__, __e__, __s__)+
7450 +ctf_array_nowrite(__t__, __n__, __e__, __s__)+
7452 +ctf_user_array(__t__, __n__, __e__, __s__)+
7454 +ctf_user_array_nowrite(__t__, __n__, __e__, __s__)+
7456 Statically-sized array of integers.
7459 Array element C type.
7465 Argument expression.
7471 +ctf_array_bitfield(__t__, __n__, __e__, __s__)+
7473 +ctf_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7475 +ctf_user_array_bitfield(__t__, __n__, __e__, __s__)+
7477 +ctf_user_array_bitfield_nowrite(__t__, __n__, __e__, __s__)+
7479 Statically-sized array of bits.
7481 The type of +__e__+ must be an integer type. +__s__+ is the number
7482 of elements of such type in +__e__+, not the number of bits.
7485 Array element C type.
7491 Argument expression.
7497 +ctf_array_text(__t__, __n__, __e__, __s__)+
7499 +ctf_array_text_nowrite(__t__, __n__, __e__, __s__)+
7501 +ctf_user_array_text(__t__, __n__, __e__, __s__)+
7503 +ctf_user_array_text_nowrite(__t__, __n__, __e__, __s__)+
7505 Statically-sized array, printed as text.
7507 The string does not need to be null-terminated.
7510 Array element C type (always `char`).
7516 Argument expression.
7522 +ctf_sequence(__t__, __n__, __e__, __T__, __E__)+
7524 +ctf_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7526 +ctf_user_sequence(__t__, __n__, __e__, __T__, __E__)+
7528 +ctf_user_sequence_nowrite(__t__, __n__, __e__, __T__, __E__)+
7530 Dynamically-sized array of integers.
7532 The type of +__E__+ must be unsigned.
7535 Array element C type.
7541 Argument expression.
7544 Length expression C type.
7550 +ctf_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7552 +ctf_user_sequence_hex(__t__, __n__, __e__, __T__, __E__)+
7554 Dynamically-sized array of integers, displayed in base 16.
7556 The type of +__E__+ must be unsigned.
7559 Array element C type.
7565 Argument expression.
7568 Length expression C type.
7573 |+ctf_sequence_network(__t__, __n__, __e__, __T__, __E__)+
7575 Dynamically-sized array of integers in network byte order (big-endian),
7576 displayed in base 10.
7578 The type of +__E__+ must be unsigned.
7581 Array element C type.
7587 Argument expression.
7590 Length expression C type.
7596 +ctf_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7598 +ctf_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7600 +ctf_user_sequence_bitfield(__t__, __n__, __e__, __T__, __E__)+
7602 +ctf_user_sequence_bitfield_nowrite(__t__, __n__, __e__, __T__, __E__)+
7604 Dynamically-sized array of bits.
7606 The type of +__e__+ must be an integer type. +__s__+ is the number
7607 of elements of such type in +__e__+, not the number of bits.
7609 The type of +__E__+ must be unsigned.
7612 Array element C type.
7618 Argument expression.
7621 Length expression C type.
7627 +ctf_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7629 +ctf_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7631 +ctf_user_sequence_text(__t__, __n__, __e__, __T__, __E__)+
7633 +ctf_user_sequence_text_nowrite(__t__, __n__, __e__, __T__, __E__)+
7635 Dynamically-sized array, displayed as text.
7637 The string does not need to be null-terminated.
7639 The type of +__E__+ must be unsigned.
7641 The behaviour is undefined if +__e__+ is `NULL`.
7644 Sequence element C type (always `char`).
7650 Argument expression.
7653 Length expression C type.
7659 Use the `_user` versions when the argument expression, `e`, is
7660 a user space address. In the cases of `ctf_user_integer*()` and
7661 `ctf_user_float*()`, `&e` must be a user space address, thus `e` must
7664 The `_nowrite` versions omit themselves from the session trace, but are
7665 otherwise identical. This means the `_nowrite` fields won't be written
7666 in the recorded trace. Their primary purpose is to make some
7667 of the event context available to the
7668 <<enabling-disabling-events,event filters>> without having to
7669 commit the data to sub-buffers.
7675 Terms related to LTTng and to tracing in general:
7678 The http://diamon.org/babeltrace[Babeltrace] project, which includes
7679 the cmd:babeltrace command, some libraries, and Python bindings.
7681 <<channel-buffering-schemes,buffering scheme>>::
7682 A layout of sub-buffers applied to a given channel.
7684 <<channel,channel>>::
7685 An entity which is responsible for a set of ring buffers.
7687 <<event,Event rules>> are always attached to a specific channel.
7690 A reference of time for a tracer.
7692 <<lttng-consumerd,consumer daemon>>::
7693 A process which is responsible for consuming the full sub-buffers
7694 and write them to a file system or send them over the network.
7696 <<channel-overwrite-mode-vs-discard-mode,discard mode>>:: The event loss
7697 mode in which the tracer _discards_ new event records when there's no
7698 sub-buffer space left to store them.
7701 The consequence of the execution of an instrumentation
7702 point, like a tracepoint that you manually place in some source code,
7703 or a Linux kernel KProbe.
7705 An event is said to _occur_ at a specific time. Different actions can
7706 be taken upon the occurrence of an event, like record the event's payload
7709 <<channel-overwrite-mode-vs-discard-mode,event loss mode>>::
7710 The mechanism by which event records of a given channel are lost
7711 (not recorded) when there is no sub-buffer space left to store them.
7713 [[def-event-name]]event name::
7714 The name of an event, which is also the name of the event record.
7715 This is also called the _instrumentation point name_.
7718 A record, in a trace, of the payload of an event which occured.
7720 <<event,event rule>>::
7721 Set of conditions which must be satisfied for one or more occuring
7722 events to be recorded.
7724 `java.util.logging`::
7726 https://docs.oracle.com/javase/7/docs/api/java/util/logging/package-summary.html[core logging facilities].
7728 <<instrumenting,instrumentation>>::
7729 The use of LTTng probes to make a piece of software traceable.
7731 instrumentation point::
7732 A point in the execution path of a piece of software that, when
7733 reached by this execution, can emit an event.
7735 instrumentation point name::
7736 See _<<def-event-name,event name>>_.
7739 A http://logging.apache.org/log4j/1.2/[logging library] for Java
7740 developed by the Apache Software Foundation.
7743 Level of severity of a log statement or user space
7744 instrumentation point.
7747 The _Linux Trace Toolkit: next generation_ project.
7749 <<lttng-cli,cmd:lttng>>::
7750 A command-line tool provided by the LTTng-tools project which you
7751 can use to send and receive control messages to and from a
7755 The https://github.com/lttng/lttng-analyses[LTTng analyses] project,
7756 which is a set of analyzing programs that are used to obtain a
7757 higher level view of an LTTng trace.
7759 cmd:lttng-consumerd::
7760 The name of the consumer daemon program.
7763 A utility provided by the LTTng-tools project which can convert
7764 ring buffer files (usually
7765 <<persistent-memory-file-systems,saved on a persistent memory file system>>)
7768 LTTng Documentation::
7771 <<lttng-live,LTTng live>>::
7772 A communication protocol between the relay daemon and live viewers
7773 which makes it possible to see events "live", as they are received by
7776 <<lttng-modules,LTTng-modules>>::
7777 The https://github.com/lttng/lttng-modules[LTTng-modules] project,
7778 which contains the Linux kernel modules to make the Linux kernel
7779 instrumentation points available for LTTng tracing.
7782 The name of the relay daemon program.
7784 cmd:lttng-sessiond::
7785 The name of the session daemon program.
7788 The https://github.com/lttng/lttng-tools[LTTng-tools] project, which
7789 contains the various programs and libraries used to
7790 <<controlling-tracing,control tracing>>.
7792 <<lttng-ust,LTTng-UST>>::
7793 The https://github.com/lttng/lttng-ust[LTTng-UST] project, which
7794 contains libraries to instrument user applications.
7796 <<lttng-ust-agents,LTTng-UST Java agent>>::
7797 A Java package provided by the LTTng-UST project to allow the
7798 LTTng instrumentation of `java.util.logging` and Apache log4j 1.2
7801 <<lttng-ust-agents,LTTng-UST Python agent>>::
7802 A Python package provided by the LTTng-UST project to allow the
7803 LTTng instrumentation of Python logging statements.
7805 <<channel-overwrite-mode-vs-discard-mode,overwrite mode>>::
7806 The event loss mode in which new event records overwrite older
7807 event records when there's no sub-buffer space left to store them.
7809 <<channel-buffering-schemes,per-process buffering>>::
7810 A buffering scheme in which each instrumented process has its own
7811 sub-buffers for a given user space channel.
7813 <<channel-buffering-schemes,per-user buffering>>::
7814 A buffering scheme in which all the processes of a Unix user share the
7815 same sub-buffer for a given user space channel.
7817 <<lttng-relayd,relay daemon>>::
7818 A process which is responsible for receiving the trace data sent by
7819 a distant consumer daemon.
7822 A set of sub-buffers.
7824 <<lttng-sessiond,session daemon>>::
7825 A process which receives control commands from you and orchestrates
7826 the tracers and various LTTng daemons.
7828 <<taking-a-snapshot,snapshot>>::
7829 A copy of the current data of all the sub-buffers of a given tracing
7830 session, saved as trace files.
7833 One part of an LTTng ring buffer which contains event records.
7836 The time information attached to an event when it is emitted.
7839 A set of files which are the concatenations of one or more
7840 flushed sub-buffers.
7843 The action of recording the events emitted by an application
7844 or by a system, or to initiate such recording by controlling
7848 The http://tracecompass.org[Trace Compass] project and application.
7851 An instrumentation point using the tracepoint mechanism of the Linux
7852 kernel or of LTTng-UST.
7854 tracepoint definition::
7855 The definition of a single tracepoint.
7858 The name of a tracepoint.
7860 tracepoint provider::
7861 A set of functions providing tracepoints to an instrumented user
7864 Not to be confused with a _tracepoint provider package_: many tracepoint
7865 providers can exist within a tracepoint provider package.
7867 tracepoint provider package::
7868 One or more tracepoint providers compiled as an object file or as
7872 A software which records emitted events.
7874 <<domain,tracing domain>>::
7875 A namespace for event sources.
7877 <<tracing-group,tracing group>>::
7878 The Unix group in which a Unix user can be to be allowed to trace the
7881 <<tracing-session,tracing session>>::
7882 A stateful dialogue between you and a <<lttng-sessiond,session
7886 An application running in user space, as opposed to a Linux kernel
7887 module, for example.