1BABELTRACE2-INTRO(7) Babeltrace 2 manual BABELTRACE2-INTRO(7)
2
6 babeltrace2-intro - Introduction to Babeltrace 2
7
9 This manual page is an introduction to the Babeltrace 2 project.
10 The “WHAT IS BABELTRACE 2?” section describes the parts of the project
12 and shows the major changes from Babeltrace 1 to Babeltrace 2 while the
13 “BABELTRACE 2 CONCEPTS” section defines the core concepts of
14 Babeltrace 2.
15 The “TRACE PROCESSING GRAPH REPRESENTATION” section shows how some
17 concepts are visually represented in other Babeltrace 2 manual pages.
18
20 Babeltrace 2 is an open-source software project of which the purpose is
21 to process or convert traces (see
22 <https://en.wikipedia.org/wiki/Tracing_(software)>).
23 The Babeltrace 2 project includes the following parts:
25 Babeltrace 2 library (libbabeltrace2)
27 A shared library with a C API.
28 With libbabeltrace2, you can programmatically create plugins and
30 component classes, build and run trace processing graphs, and more
31 (see the “BABELTRACE 2 CONCEPTS” section for more details about
32 those concepts).
33 All the other Babeltrace 2 parts rely on this library.
35 babeltrace2 command-line program
37 A command-line interface which uses libbabeltrace2 to load plugins,
38 create a trace processing graph, create components, connect their
39 ports correctly, and run the graph.
40 You can also use babeltrace2 to list the available plugins or to
42 query an object from a component class.
43 See babeltrace2(1).
45 Babeltrace 2 Python bindings
47 A Python 3 package (bt2) which offers a Pythonic interface of
48 libbabeltrace2.
49 You can perform the same operations which are available in
51 libbabeltrace2 with the Python bindings, but more conveniently and
52 with less code. However, the Python bindings are less performant
53 than libbabeltrace2.
54 Babeltrace 2 project’s plugins
56 The Babeltrace 2 plugins shipped with the project.
57 Those plugins are not special in that they only rely on
59 libbabeltrace2 and you don’t need them to use libbabeltrace2,
60 babeltrace2(1), or the Python bindings. However, the project’s
61 plugins provide many widely used trace format encoders/decoders as
62 well as common trace processing graph utilities.
63 The Babeltrace 2 project’s plugins are:
65 ctf
67 Common Trace Format (see <https://diamon.org/ctf/>) (CTF)
68 input/output, including the LTTng live source.
69 See babeltrace2-plugin-ctf(7).
71 lttng-utils
73 Graph utilities specific to LTTng (see <https://lttng.org/>)
74 traces.
75 See babeltrace2-plugin-lttng-utils(7).
77 text
79 Plain text input/output.
80 See babeltrace2-plugin-text(7).
82 utils
84 Common graph utilities (muxer, trimmer, counter, dummy sink).
85 See babeltrace2-plugin-utils(7).
87 Changes since Babeltrace 1
89 This manual page is an introduction to Babeltrace 2, a rewrite of
90 Babeltrace 1 with a focus on extensibility, flexibility, and
91 interoperability.
92 Babeltrace 1 exists since 2010.
94 You can install both projects on the same file system as there are no
96 file name conflicts.
97 The major improvements brought by Babeltrace 2 are:
99 General
101 • Full plugin support: any user can distribute a Babeltrace 2
103 plugin and, as long as libbabeltrace2 finds it, any application
104 linked to libbabeltrace2 can load it and use it.
105 Plugins are not just trace format encoders and decoders: they
107 package source, filter, and sink component classes so that you
108 can connect specialized, reusable components together in a
109 trace processing graph to create a customized trace conversion
110 or analysis device.
111 This modular strategy is much like how the FFmpeg (see
113 <https://www.ffmpeg.org/>), GStreamer (see
114 <https://gstreamer.freedesktop.org/>), and DirectShow (see
115 <https://en.wikipedia.org/wiki/DirectShow>) projects approach
116 media stream processing.
117 • All the parts of the Babeltrace 2 project run on the major
119 operating systems, including Windows and macOS.
120 • Some component classes, such as sink.text.pretty (similar to
122 the text output format of babeltrace(1)) and sink.text.details,
123 can write color codes to the standard output when it’s
124 connected to a color-enabled terminal.
125 The Babeltrace 2 log, printed to the standard output, can also
127 be colorized.
128 Command-line interface
130 • Whereas you can convert traces from one format to another with
132 Babeltrace 1’s CLI tool, babeltrace(1), you can also execute a
133 custom trace manipulation task with babeltrace2(1) thanks to
134 the babeltrace2-run(1) command.
135 • The babeltrace2-convert(1) command features an automatic source
137 component discovery algorithm to find the best suited
138 components to create for a given non-option argument (file or
139 directory path, or custom string like an LTTng live (see
140 <https://lttng.org>) URL).
141 For example:
143 $ babeltrace2 /path/to/ctf/trace
145 $ babeltrace2 net://localhost/host/myhost/my-session
147 CTF (see <https://diamon.org/ctf/>) input/output
149 • The source.ctf.fs component class, which is more or less the
151 equivalent of Babeltrace 1’s ctf input format, has features not
152 found in Babeltrace 1:
153 • The component handles many trace quirks which are the
155 results of known tracer bugs and corner cases (LTTng-UST,
156 LTTng-modules, and barectf (see
157 <https://github.com/efficios/barectf>)), making it possible
158 to decode malformed packets.
159 • The component merges CTF traces sharing the same UUID into
161 a single, logical trace.
162 This feature supports LTTng 2.11’s tracing session rotation
164 trace chunks.
165 • With a sink.ctf.fs component, you can create CTF traces on the
167 file system.
168 With babeltrace2(1), you can use the --output-format=ctf and
170 --output options to create an implicit sink.ctf.fs component.
171 For example:
173 $ babeltrace2 /path/to/input/trace \
175 --output-format=ctf --output=trace-dir
176 LTTng live (see <https://lttng.org>) input
178 • The babeltrace(1) command exits successfully when it cannot
180 find an LTTng live (--input-format=lttng-live option) tracing
181 session.
182 The session-not-found-action initialization parameter controls
184 what a source.ctf.lttng-live message iterator does when it
185 cannot find the remote tracing session.
186 If the action is end, the message iterator does like
188 babeltrace(1) and simply ends successfully.
189 If the action is continue (the default), the message iterator
191 never ends: it keeps on trying until the tracing session
192 exists, indeed subscribing to the session.
193 Library
195 • libbabeltrace2 shares nothing with libbabeltrace.
197 The Babeltrace 2 library C API has features such as:
199 • A single header file.
201 • Function precondition and postcondition checking.
203 • Object-oriented model with shared and unique objects.
205 • Strict C typing and const correctness.
207 • User-extensible classes.
209 • Rich, thread-safe error reporting.
211 • Per-component and per-subsystem logging levels.
213 • Trace intermediate representation (IR) objects to make the
215 API trace-format-agnostic.
216 • A versioned protocol for message interchange between
218 components to enable forward and backward compatibility.
219 • You can build the library in developer mode to enable an
221 extensive set of function precondition and postcondition
222 checks.
223 The developer mode can help detect programming errors early
225 when you develop a Babeltrace 2 plugin or an application using
226 libbabeltrace2.
227 See the project’s README for build-time requirements and
229 detailed build instructions.
230
232 This section defines the main concepts of the Babeltrace 2 project.
233 These concepts translate into types and functions in libbabeltrace2 and
235 its Python bindings, but also as command-line actions and options in
236 the babeltrace2 program. The other Babeltrace 2 manual pages assume
237 that you are familiar with the following definitions.
238 Some Babeltrace 2 concepts are interdependent: it is normal to jump
240 from one definition to another to understand the big picture.
241 Component class
243 A reusable class which you can instantiate as one or more
244 components within a trace processing graph.
245 There are three types of component classes used to create the three
247 types of components: source, filter, and sink.
248 A component class implements methods, one of which is an
250 initialization method, or constructor, to create a component. You
251 pass initialization parameters to this method to customize the
252 created component. For example, the initialization method of the
253 source.ctf.fs component class accepts a mandatory inputs parameter
254 which is an array of file system path(s) to the CTF trace(s). It
255 also accepts an optional clock-class-offset-ns parameter which is
256 an offset, in nanoseconds, to add to all the clock classes
257 (descriptors of stream clocks) found in the traces’s metadata.
258 A component class can have a description and a help text.
260 Component
262 A node within a trace processing graph.
263 There are three types of components:
265 Source component
267 An input component which produces messages.
268 Examples: CTF files input, log file input, LTTng live input,
270 random event generator.
271 Filter component
273 An intermediate component which can transform the messages it
274 consumes, augment them, sort them, discard them, or create new
275 ones.
276 Examples: filter which removes messages based on an expression,
278 filter which adds debugging information to selected events,
279 message muxer, trace trimmer.
280 Sink component
282 An output component which consumes messages and usually writes
283 them to one or more formatted files.
284 Examples: log file output, CTF files output, pretty-printed
286 plain text output.
287 Components are connected together within a trace processing graph
289 through their ports. Source components have output ports, sink
290 components have input ports, and filter components have both.
291 A component is the instance of a component class. The terms
293 component and component class instance are equivalent.
294 Within a trace processing graph, each component has a unique name.
296 This is not the name of its component class, but an instance name.
297 If human is a component class name, than Nancy and John could be
298 component names.
299 Once a graph is configured (the first time it runs), you cannot add
301 components to it for the remaining graph’s lifetime.
302 Port
304 A connection point, on a component, from which are sent or where
305 are received messages when the trace processing graph runs.
306 An output port is from where messages are sent. An input port is
308 where messages are received. Source components have output ports,
309 sink components have input ports, and filter components have both.
310 You can only connect an output port to a single input port.
312 All ports do not need to be connected.
314 A filter or sink component receiving messages from its input ports
316 is said to consume messages.
317 The link between an output port and input port is a connection.
319 Once a graph is configured (the first time it runs), you cannot
321 connect ports for the remaining graph’s lifetime.
322 Connection
324 The link between an output port and an input port through which
325 messages flow when a trace processing graph runs.
326 Message iterator
328 An iterator on an input port of which the returned elements are
329 messages.
330 A component or another message iterator can create many message
332 iterators on a single input port, before or while the trace
333 processing graph runs.
334 Message
336 The element of a message iterator.
337 Messages flow from output ports to input ports.
339 A source component message iterator produces messages, while a sink
341 component consumes them. A filter component message iterator can
342 both consume and produce messages.
343 The main types of messages are:
345 Event
347 A trace event record within a packet or within a stream.
348 Packet beginning
350 The beginning of a packet within a stream.
351 A packet is a conceptual container of events.
353 Packet end
355 The end of a packet within a stream.
356 Stream beginning
358 The beginning of a stream.
359 A stream is a conceptual container of packets and/or events.
361 Usually, a given source component’s output port sends packet
363 and event messages which belong to a single stream, but it’s
364 not required.
365 Stream end
367 The end of a stream.
368 Discarded events
370 A count of discarded events within a given time interval for a
371 given stream.
372 Discarded packets
374 A count of discarded packets within a given time interval for a
375 given stream.
376 Trace processing graph
378 A filter graph (see <https://en.wikipedia.org/wiki/Filter_graph>)
379 where nodes are components and messages flow from output ports to
380 input ports.
381 You can build a trace processing graph with libbabeltrace2, with
383 the Babeltrace 2 Python bindings, or with the babeltrace2-run(1)
384 and babeltrace2-convert(1) CLI commands.
385 When a trace processing graph runs, the sink components consume
387 messages from their input ports, making all the graph’s message
388 iterators work one message at a time to perform the trace
389 conversion or analysis duty.
390 Plugin
392 A container, or package, of component classes as a shared library
393 or Python module.
394 Each component class within a plugin has a type (source, filter, or
396 sink) and a name. The type and name pair is unique within a given
397 plugin.
398 libbabeltrace2 can load a plugin (.so, .dll, or .py file) at run
400 time: the result is a plugin object in which you can find a
401 specific component class and instantiate it within a trace
402 processing graph as a component.
403 The babeltrace2 program uses the
405 COMP-CLS-TYPE.PLUGIN-NAME.COMP-CLS-NAME format to identify a
406 specific component class within a specific plugin. COMP-CLS-TYPE
407 is either source (or src), filter (or flt), or sink.
408 You can list the available Babeltrace 2 plugins with the
410 babeltrace2-list-plugins(1) command.
411 Query
413 An operation with which you can get a named object from a component
414 class, possibly with custom query parameters.
415 The plain text metadata stream of a CTF trace and the available
417 LTTng live sessions of a given LTTng relay daemon are examples of
418 query objects.
419 You can use libbabeltrace2, the Babeltrace 2 Python bindings, or
421 the babeltrace2-query(1) CLI command to query a component class’s
422 object.
423
425 In the Babeltrace 2 manual pages, a component is represented with a
426 box. The box has the component class type, plugin name, and component
427 class name at the top. Just below, between square brackets, is its
428 component name within the trace processing graph. Each port is
429 represented with an @ symbol on the border(s) of the component box with
430 its name inside the box. Output ports are on the box’s right border
431 while input ports are on the box’s left border.
432 For example, here’s a source component box:
434 +------------+
436 | src.ctf.fs |
437 | [my-src] |
438 | |
439 | stream0 @
440 | stream1 @
441 | stream2 @
442 +------------+
443 This one is an instance of the source.ctf.fs component class named my-
445 src. It has three output ports named stream0, stream1, and stream2.
446 A trace processing graph is represented with multiple component boxes
448 connected together. The connections are arrows from output ports to
449 input ports.
450 For example, here’s a simple conversion graph:
452 +------------+ +-----------------+ +------------------+
454 | src.ctf.fs | | flt.utils.muxer | | sink.text.pretty |
455 | [ctf] | | [muxer] | | [text] |
456 | | | | | |
457 | stream0 @--->@ in0 out @--->@ in |
458 | stream1 @--->@ in1 | +------------------+
459 | stream2 @--->@ in2 |
460 +------------+ @ in3 |
461 +-----------------+
462 Note that input port in3 of component muxer is not connected in this
464 example.
465 Sometimes, we symbolically represent other resources which are consumed
467 from or produced by components. In this case, arrows are used, but they
468 do not go to or from port symbols (@), except for messages. For
469 example, in the graph above, the ctf source component consumes a CTF
470 trace and the text sink component prints plain text to the terminal, so
471 here’s a more complete diagram:
472 CTF trace
474 |
475 | +------------+ +-----------------+ +------------------+
476 | | src.ctf.fs | | flt.utils.muxer | | sink.text.pretty |
477 '-->| [ctf] | | [muxer] | | [text] |
478 | | | | | |
479 | stream0 @--->@ in0 out @--->@ in |
480 | stream1 @--->@ in1 | +-----+------------+
481 | stream2 @--->@ in2 | |
482 +------------+ @ in3 | '--> Terminal
483 +-----------------+
484 Here’s another example of a more complex graph which splits a specific
486 stream using some criteria:
487 +------------+ +-----------------+ +------------------+
489 | src.ctf.fs | | flt.utils.muxer | | sink.text.pretty |
490 | [ctf-in] | | [muxer] | | [text] |
491 | | | | | |
492 | stream0 @--->@ in0 out @--->@ in |
493 | stream1 @--->@ in1 | +------------------+
494 | stream2 @-. @ in2 |
495 +------------+ | +-----------------+ +-------------+
496 | | sink.ctf.fs |
497 | | [ctf-out0] |
498 | +-------------------+ | |
499 | | flt.some.splitter | .->@ in |
500 | | [splitter] | | +-------------+
501 | | | |
502 '->@ in A @-' +-------------+
503 | B @-. | sink.ctf.fs |
504 +-------------------+ | | [ctf-out1] |
505 | | |
506 '->@ in |
507 +-------------+
508
510 If you encounter any issue or usability problem, please report it on
511 the Babeltrace bug tracker (see
512 <https://bugs.lttng.org/projects/babeltrace>).
513
515 The Babeltrace project shares some communication channels with the
516 LTTng project (see <https://lttng.org/>).
517 • Babeltrace website (see <https://babeltrace.org/>)
519 • Mailing list (see <https://lists.lttng.org>) for support and
521 development: lttng-dev@lists.lttng.org
522 • IRC channel (see <irc://irc.oftc.net/lttng>): #lttng on
524 irc.oftc.net
525 • Bug tracker (see <https://bugs.lttng.org/projects/babeltrace>)
527 • Git repository (see <https://git.efficios.com/?p=babeltrace.git>)
529 • GitHub project (see <https://github.com/efficios/babeltrace>)
531 • Continuous integration (see
533 <https://ci.lttng.org/view/Babeltrace/>)
534 • Code review (see <https://review.lttng.org/q/project:babeltrace>)
536
538 The Babeltrace 2 project is the result of hard work by many regular
539 developers and occasional contributors.
540 The current project maintainer is Jérémie Galarneau
542 <mailto:jeremie.galarneau@efficios.com>.
543
545 This manual page is part of the Babeltrace 2 project.
546 Babeltrace is distributed under the MIT license (see
548 <https://opensource.org/licenses/MIT>).
549
551 babeltrace2(1)
552Babeltrace 2.0.4 14 September 2019 BABELTRACE2-INTRO(7)