1 /*! \page faq Frequently Asked Questions
3 \htmlinclude .FAQ.doc.toc
5 \section faq_installation Installing the SimGrid library
7 Many people have been asking me questions on how to use SimGrid. Quite
8 often, the questions were not really about SimGrid but on the
9 installation process. This section is intended to help people that are
10 not familiar with compiling C files under UNIX. If you follow these
11 instructions and still have some troubles, drop an e-mail to
12 <simgrid-user@lists.gforge.inria.fr>.
14 \subsection faq_compiling Compiling SimGrid from an archive
16 First of all, you need to download the latest version of SimGrid from
17 <a href="http://gforge.inria.fr/frs/?group_id=12">here</a>.
18 Suppose you have uncompressed SimGrid in some temporary location of
19 your home directory (say <tt>/home/joe/tmp/simgrid-3.0.1 </tt>). The
20 simplest way to use SimGrid is to install it in your home
21 directory. Change your directory to
22 <tt>/home/joe/tmp/simgrid-3.0.1</tt> and type
24 \verbatim./configure --prefix=$HOME
29 If at some point, something fails, check the section "\ref
30 faq_compil_trouble". If it does not help, you can report this problem to the
31 list but, please, avoid sending a laconic mail like "There is a problem. Is it
32 okay?". Send the config.log file which is automatically generated by
33 configure. Try to capture both the standard output and the error output of the
34 <tt>make</tt> command with <tt>script</tt>. There is no way for us to help you
35 without the relevant bits of information.
37 Now, the following directory should have been created :
39 \li <tt>/home/joe/doc/simgrid/html/</tt>
40 \li <tt>/home/joe/lib/</tt>
41 \li <tt>/home/joe/include/</tt>
43 SimGrid is not a binary, it is a library. Both a static and a dynamic
44 version are available. Here is what you can find if you try a <tt>ls
47 \verbatim libsimgrid.a libsimgrid.la libsimgrid.so libsimgrid.so.0 libsimgrid.so.0.0.1
50 Thus, there is two ways to link your program with SimGrid:
51 \li Either you use the static version, e.g
52 \verbatim gcc libsimgrid.a -o MainProgram MainProgram.c
54 In this case, all the SimGrid functions are directly
55 included in <tt>MainProgram</tt> (hence a bigger binary).
56 \li Either you use the dynamic version (the preferred method)
57 \verbatim gcc -lsimgrid -o MainProgram MainProgram.c
59 In this case, the SimGrid functions are not included in
60 <tt>MainProgram</tt> and you need to set your environment
61 variable in such a way that <tt>libsimgrid.so</tt> will be
62 found at runtime. This can be done by adding the following
63 line in your .bashrc (if you use bash and if you have
64 installed the SimGrid libraries in your home directory):
65 \verbatim export LD_LIBRARY_PATH=$HOME/lib/:$LD_LIBRARY_PATH
69 \subsection faq_compiling_cvs Compiling SimGrid from the CVS
71 The project development takes place in the cvs, where all changes are
72 commited when they happen. Then every once in a while, we make sure that the
73 code quality meets our standard and release an archive from the code in the
74 CVS. We afterward go back to the development in the CVS. So, if you need a
75 recently added feature and can afford some little problem with the stability
76 of the lastest features, you may want to use the CVS version instead of a
79 For that, you first need to get the "simgrid" module from
80 <a href="http://gforge.inria.fr/scm/?group_id=12">here</a>.
82 You won't find any <tt>configure</tt> and a few other things
83 (<tt>Makefile.in</tt>'s, documentation, ...) will be missing as well. The
84 reason for that is that all these files have to be regenerated using the
85 latest versions of <tt>autoconf</tt>, <tt>libtool</tt>, <tt>automake</tt>
86 (>1.9) and <tt>doxygen</tt> (>1.4). To generate the <tt>configure</tt> and
87 the <tt>Makefile.in</tt>'s, you just have to launch the <tt>bootstrap</tt>
88 command that resides in the top of the source tree. Then just follow the
89 instructions of Section \ref faq_compiling.
91 We insist on the fact that you really need the latest versions of
92 autoconf and automake. Doing this step on exotic architectures/systems
93 (i.e. anything different from a recent linux distribution) may be
94 ... uncertain. If you want to use the CVS version on another
95 architecture/system, you should do the previous steps on a perfectly
96 standard box, then do a <tt>make dist</tt> that will build you a
97 perfectly portable SimGrid archive.
99 In summary, the following commands will checkout the CVS, regenerate the
100 configure script and friends, configure SimGrid and build an archive you can
101 use on another machine afterward.
103 \verbatim cvs -d :pserver:anonymous@scm.gforge.inria.fr:/cvsroot/simgrid login
104 cvs -d :pserver:anonymous@scm.gforge.inria.fr:/cvsroot/simgrid checkout simgrid
107 ./configure --enable-maintainer-mode
108 make dist \endverbatim
110 \subsection faq_setting_MSG Setting up your own MSG code
112 Do not build your simulator by modifying the SimGrid examples. Go
113 outside the SimGrid source tree and create your own working directory
114 (say <tt>/home/joe/SimGrid/MyFirstScheduler/</tt>).
116 Suppose your simulation has the following structure (remember it is
117 just an example to illustrate a possible way to compile everything;
118 feel free to organize it as you want).
120 \li <tt>sched.h</tt>: a description of the core of the
121 scheduler (i.e. which functions are can be used by the
122 agents). For example we could find the following functions
123 (master, forwarder, slave).
125 \li <tt>sched.c</tt>: a C file including <tt>sched.h</tt> and
126 implementing the core of the scheduler. Most of these
127 functions use the MSG functions defined in section \ref
130 \li <tt>masterslave.c</tt>: a C file with the main function, i.e.
131 the MSG initialization (MSG_global_init()), the platform
132 creation (e.g. with MSG_create_environment()), the
133 deployment phase (e.g. with MSG_function_register() and
134 MSG_launch_application()) and the call to
137 To compile such a program, we suggest to use the following
138 Makefile. It is a generic Makefile that we have used many times with
139 our students when we teach the C language.
143 masterslave: masterslave.o sched.o
145 INSTALL_PATH = $$HOME
147 PEDANTIC_PARANOID_FREAK = -O0 -Wshadow -Wcast-align \
148 -Waggregate-return -Wmissing-prototypes -Wmissing-declarations \
149 -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations \
150 -Wmissing-noreturn -Wredundant-decls -Wnested-externs \
151 -Wpointer-arith -Wwrite-strings -finline-functions
152 REASONABLY_CAREFUL_DUDE = -Wall
153 NO_PRAYER_FOR_THE_WICKED = -w -O2
154 WARNINGS = $(REASONABLY_CAREFUL_DUDE)
155 CFLAGS = -g $(WARNINGS)
157 INCLUDES = -I$(INSTALL_PATH)/include
158 DEFS = -L$(INSTALL_PATH)/lib/
159 LDADD = -lm -lsimgrid
163 $(CC) $(INCLUDES) $(DEFS) $(CFLAGS) $^ $(LIBS) $(LDADD) -o $@
166 $(CC) $(INCLUDES) $(DEFS) $(CFLAGS) -c -o $@ $<
169 rm -f $(BIN_FILES) *.o *~
175 The first two lines indicates what should be build when typing make
176 (<tt>masterslave</tt>) and of which files it is to be made of
177 (<tt>masterslave.o</tt> and <tt>sched.o</tt>). This makefile assumes
178 that you have set up correctly your <tt>LD_LIBRARY_PATH</tt> variable
179 (look, there is a <tt>LDADD = -lm -lsimgrid</tt>). If you prefer using
180 the static version, remove the <tt>-lsimgrid</tt> and add a
181 <tt>$(INSTALL_PATH)/lib/libsimgrid.a</tt> on the next line, right
182 after the <tt>LIBS = </tt>.
184 More generally, if you have never written a Makefile by yourself, type
185 in a terminal : <tt>info make</tt> and read the introduction. The
186 previous example should be enough for a first try but you may want to
187 perform some more complex compilations...
189 \subsection faq_setting_GRAS Setting up your own GRAS code
191 If you use the GRAS interface instead of the MSG one, then previous section
192 is not the better source of information. Instead, you should check the GRAS
193 tutorial in general, and the \ref GRAS_tut_tour_setup in particular.
195 \section faq_simgrid I'm new to SimGrid. I have some questions. Where should I start?
197 You are at the right place... Having a look to these
198 <a href="http://graal.ens-lyon.fr/~alegrand/articles/slides_g5k_simul.pdf">slides</a>
200 <a href="http://graal.ens-lyon.fr/~alegrand/articles/Simgrid-Introduction.pdf">"obsolete" slides</a>)
201 may give you some insights on what SimGrid can help you to do and what
202 are its limitations. Then you definitely should read the \ref
203 MSG_examples. There is also a mailing list: <simgrid-user@lists.gforge.inria.fr>.
205 \subsection faq_generic Building a generic simulator
207 Please read carefully the \ref MSG_examples. You'll find in \ref
208 MSG_ex_master_slave a very simple consisting of a master (that owns a bunch of
209 tasks and distributes them) , some slaves (that process tasks whenever
210 they receive one) and some forwarder agents (that simply pass the
211 tasks they receive to some slaves).
213 \subsection faq_visualization Visualizing the schedule
215 It is sometime convenient to "see" how the agents are behaving. If you
216 like colors, you can use <tt>tools/MSG_visualization/colorize.pl </tt>
217 as a filter to your MSG outputs. It works directly with INFO. Beware,
218 INFO() prints on stderr. Do not forget to redirect if you want to
219 filter (e.g. with bash):
221 ./msg_test small_platform.xml small_deployment.xml 2>&1 | ../../tools/MSG_visualization/colorize.pl
224 We also have a more graphical output. Have a look at MSG_paje_output(). It
225 generates an input to <a href="http://www-id.imag.fr/Logiciels/paje/">Paje</a>.
228 <a href="Paje_MSG_screenshot.jpg"><img src="Paje_MSG_screenshot_thn.jpg"></a>
232 \subsection faq_postmortem_analysis Online/postmortem analysis
234 Vizualization with Paje can be seen as a kind of postmortem
235 analysis. However, as soon as you start playing with big simulations,
236 you'll realize that processing such output is kind of tricky. There is
237 so much generic informations that it is hard to find the information
240 As a matter of fact, loging really depends on simulations (e.g. what
241 kind of events is important...). That is why we do not propose a big
242 dump of your whole simulation (it would slow everything down) but give
243 you neat tools to structure you logs. Have a look at \ref XBT_log. In
244 fact, rather than a post-mortem analysis, you may want to do it on the
245 fly. The process you are running can do whatever you want. Have you
246 thought about adding a global structure where you directly compute the
247 informations that are really important rather than writing everything
248 down and then processing huge files ?
250 \subsection faq_C Argh! Do I really have to code in C ?
252 Up until now, there is no binding for other languages. If you use C++,
253 you should be able to use the SimGrid library as a standard C library
254 and everything should work fine (simply <i>link</i> against this
255 library; recompiling SimGrid with a C++ compiler won't work and it
256 wouldn't help if you could).
258 In fact, the bindings needed to allow one to use SimGrid from Perl,
259 Python, Java, etc. are double-layered. The first layer would allow
260 you to call for example the MSG_task_get_name(task) function while
261 what you really want is a proper object wrapping allowing you to call
262 task->name(). That's the purpose of the second layer. The first one
263 is granted with C++ but can be done with tools like
264 <a href="www.swig.org/">swig</a> for other languages like Perl, Ruby,
265 Python, CAML. None of us really need the second one (which is a bit
266 more demanding and cannot be automatically generated) yet and there is
267 no real point in doing the first one without the second. :)
269 As usual, you're welcome to participate.
271 \section faq_MIA How to ....? Is there a function in the API to simply ....?
273 Here is the deal. The whole SimGrid project (MSG, SURF, GRAS, ...) is
274 meant to be kept as simple and generic as possible. We cannot add
275 functions for everybody's need when these functions can easily be
276 built from the ones already in the API. Most of the time, it is
277 possible and when it was not possible we always have upgraded the API
278 accordingly. When somebody asks us a question like "How to do that ?
279 Is there a function in the API to simply do this ?", we're always glad
280 to answer and help. However if we don't need this code for our own
281 need, there is no chance we're going to write it... it's your job! :)
282 The counterpart to our answers is that once you come up with a neat
283 implementation of this feature (task duplication, RPC, thread
284 synchronization, ...), you should send it to us and we will be glad to
285 add it to the distribution. Thus, other people will take advantage of
286 it (and we don't have to answer this question again and again ;).
288 You'll find in this section a few "Missing In Action" features. Many
289 people have asked about it and we have given hints on how to simply do
290 it with MSG. Feel free to contribute...
292 \subsection faq_MIA_examples I want some more complex examples!
294 Many people have come to ask me a more complex example and each time,
295 they have realized afterward that the basics were in the previous three
298 Of course they have often been needing more complex functions like
299 MSG_process_suspend(), MSG_process_resume() and
300 MSG_process_isSuspended() (to perform synchronization), or
301 MSG_task_Iprobe() and MSG_process_sleep() (to avoid blocking
302 receptions), or even MSG_process_create() (to design asynchronous
303 communications or computations). But the examples are sufficient to
306 We know. We should add some more examples, but not really some more
307 complex ones... We should add some examples that illustrate some other
308 functionalities (like how to simply encode asynchronous
309 communications, RPC, process migrations, thread synchronization, ...)
310 and we will do it when we will have a little bit more time. We have
311 tried to document the examples so that they are understandable. Tell
312 us if something is not clear and once again feel free to participate!
315 \subsection faq_MIA_taskdup Missing in action: Task duplication/replication
317 There is no task duplication in MSG. When you create a task, you can
318 process it or send it somewhere else. As soon as a process has sent
319 this task, he doesn't have this task anymore. It's gone. The receiver
320 process has got the task. However, you could decide upon receiving to
321 create a "copy" of a task but you have to handle by yourself the
322 semantic associated to this "duplication".
324 As we already told, we prefer keeping the API as simple as
325 possible. This kind of feature is rather easy to implement by users
326 and the semantic you associate really depends on people. Having a
327 *generic* task duplication mechanism is not that trivial (in
328 particular because of the data field). That is why I would recommand
329 that you write it by yourself even if I can give you advice on how to
332 You have the following functions to get informations about a task:
333 MSG_task_get_name(), MSG_task_get_compute_duration(),
334 MSG_task_get_remaining_computation(), MSG_task_get_data_size(),
335 and MSG_task_get_data().
337 You could use a dictionnary (#xbt_dict_t) of dynars (#xbt_dynar_t). If
338 you still don't see how to do it, please come back to us...
340 \subsection faq_MIA_asynchronous I want to do asynchronous communications in MSG
342 Up until now, there is no asynchronous communications in MSG. However,
343 you can create as many process as you want so you should be able to do
344 whatever you want... I've written a queue module to help implementing
345 some asynchronous communications at low cost (creating thousands of
346 process only to handle communications may be problematic in term of
347 performance at some point). I'll add it in the distribution asap.
349 \subsection faq_MIA_thread_synchronization I need to synchronize my MSG processes
351 You obviously cannot use pthread_mutexes of pthread_conds. The best
352 thing would be to propose similar structures. Unfortunately, we
353 haven't found time to do it yet. However you can try to play with
354 MSG_process_suspend() and MSG_process_resume(). You can even do some
355 synchronization with fake communications (using MSG_task_get(),
356 MSG_task_put() and MSG_task_Iprobe()).
358 \subsection faq_MIA_host_load Where is the get_host_load function hidden in MSG?
360 There is no such thing because its semantic wouldn't be really
361 clear. Of course, it is something about the amount of host throughput,
362 but there is as many definition of "host load" as people asking for
363 this function. First, you have to remember that resource availability
364 may vary over time, which make any load notion harder to define.
366 It may be instantaneous value or an average one. Moreover it may be only the
367 power of the computer, or may take the background load into account, or may
368 even take the currently running tasks into account. In some SURF models,
369 communications have an influence on computational power. Should it be taken
372 So, we decided not to include such a function into MSG and let people do it
373 thereselves so that they get the value matching exactly what they mean. One
374 possibility is to run active measurement as in next code snippet. It is very
375 close from what you would have to do out of the simulator, and thus gives
376 you information that you could also get in real settings to not hinder the
377 realism of your simulation.
380 double get_host_load() {
381 m_task_t task = MSG_task_create("test", 0.001, 0, NULL);
382 double date = MSG_get_clock();
384 MSG_task_execute(task);
385 date = MSG_get_clock() - date;
386 MSG_task_destroy(task);
391 Of course, it may not match your personal definition of "host load". In this
392 case, please detail what you mean on the mailing list, and we will extend
393 this FAQ section to fit your taste if possible.
395 \subsection faq_MIA_batch_scheduler Is there a native support for batch schedulers in SimGrid ?
397 No, there is no native support for batch schedulers and none is
398 planned because this is a very specific need (and doing it in a
399 generic way is thus very hard). However some people have implemented
400 their own batch schedulers. Vincent Garonne wrote one during his PhD
401 and put his code in the contrib directory of our CVS so that other can
402 keep working on it. You may find inspinring ideas in it.
404 \subsection faq_MIA_checkpointing I need a checkpointing thing
406 Actually, it depends on whether you want to checkpoint the simulation, or to
407 simulate checkpoints.
409 The first one could help if your simulation is a long standing process you
410 want to keep running even on hardware issues. It could also help to
411 <i>rewind</i> the simulation by jumping sometimes on an old checkpoint to
412 cancel recent calculations.\n
413 Unfortunately, such thing will probably never exist in SG. One would have to
414 duplicate all data structures because doing a rewind at the simulator level
415 is very very hard (not talking about the malloc free operations that might
416 have been done in between). Instead, you may be interested in the Libckpt
417 library (http://www.cs.utk.edu/~plank/plank/www/libckpt.html). This is the
418 checkpointing solution used in the condor project, for example. It makes it
419 easy to create checkpoints (at the OS level, creating something like core
420 files), and rerunning them on need.
422 If you want to simulate checkpoints instead, it means that you want the
423 state of an executing task (in particular, the progress made towards
424 completion) to be saved somewhere. So if a host (and the task executing on
425 it) fails (cf. #MSG_HOST_FAILURE), then the task can be restarted
426 from the last checkpoint.\n
428 Actually, such a thing does not exists in SimGrid either, but it's just
429 because we don't think it is fundamental and it may be done in the user code
430 at relatively low cost. You could for example use a watcher that
431 periodically get the remaining amount of things to do (using
432 MSG_task_get_remaining_computation()), or fragment the task in smaller
435 \section faq_SG Where has SG disappeared?!?
437 OK, it's time to explain what's happening to the SimGrid project. Let's
438 start with a little bit of history.
440 * Historically, SimGrid was a low-level toolkit for scheduling with
441 classical models such as DAGs. That was SimGrid v.1.* aka SG, written
442 by Henri Casanova. I (Arnaud) had been using it in its earliest
443 versions during an internship at UCSD.
445 Then we have realized that encoding distributed algorithm in SG was a
448 * So we have built MSG on top of SG and have released SimGrid v.2.*. MSG
449 offered a very basic API to encode a distributed application easily.
450 However encoding MSG on top of SG was not really convenient and did not
451 use the DAG part since the control of the task synchronization was done
452 on top of MSG and no more in SG. We have been playing a little bit with
453 MSG. We have realized that:
455 \li 1) the platform modeling was quite flexible and could be "almost"
456 automated (e.g. using random generator and post-annotations);
458 \li 2) SG was the bottleneck because of the way we were using
459 it. We needed to simulate concurrent transfers, complex load
460 sharing mechanisms. Many optimizations (e.g. trace integration)
461 were totally inefficient when combined with MSG and made extending SG
462 to implement new sharing policies, parallel tasks models, or failures
463 (many people were asking for these kind of features) a real pain;
465 \li 3) the application modeling was not really easy. Even though the
466 application modeling depends on people's applications, we thought
467 we could improve things here. One of our target here was realistic
468 distributed applications ranging from computer sensor networks like
469 the NWS to peer-to-peer applications;
471 * So we have been planning mainly two things for SimGrid 3:
473 \li 1) I have proposed to get rid of SG and to re-implement a new kernel
474 that would be faster and more flexible. That is what I did in the
475 end of 2004: SURF. SURF is based on a fast max-min linear solver
476 using O(1) data-structures. I have quickly replaced SG by SURF in
477 MSG and the result has been that on the MSG example, the new
478 version was more than 10 times faster while we had gain a lot of
479 flexibility. I think I could still easily make MSG faster but I
480 have to work on MSG now (e.g. using some of the O(1)
481 data-structures I've been using to build SURF) since it has become
482 the bottleneck. Some MSG functions have been removed from the API
483 but they were mainly intended to build the platform by hand (they
484 had appeared in the earliest versions of MSG) and were therefore
485 not useful anymore since we are providing a complete mechanism to
486 automatically build the platform and deploy the agents on it.;
488 \li 2) GRAS is a new project Martin and I have come up with. The idea is
489 to have a programming environment that let you program real
490 distributed applications while letting you the ability to run it in
491 the simulator without having to change the slightest line of your
492 code. From the simulation point of view, GRAS performs the
493 application modeling automatically... Up until now, GRAS works on
494 top MSG for historical reasons but I'm going to make it work
495 directly on top of SURF so that it can use all the flex and the
496 speed provided by SURF.
498 Those two things are working, but we want to make everything as clean as
499 possible before releasing SimGrid v.3.
501 So what about those nice DAGs we used to have in SimGrid v.1.? They're
502 not anymore in SimGrid v.3. At least not in their original form... Let
503 me recall you the way SimGrid 3 is organized:
517 XBT is our tool box and now, you should have an idea of what the other
518 ones are. As you can see, the primitive SG is not here
519 anymore. However we have written a brand new and cleaner API for this
520 purpose: \ref SD_API. It is built directly on top of SURF and provides
521 an API rather close to the old SG:
524 ______________________
526 |____________________|
527 | | MSG | GRAS | SD |
528 | -------------------|
530 | -------------------|
532 ----------------------
535 The nice thing is that, as it is writen on top of SURF, it seamlessly
536 support DAG of parallel tasks as well as complex communications
537 patterns. Some old codes using SG are currently under rewrite using
538 \ref SD_API to check that all needful functions are provided.
540 \subsection faq_SG_DAG How to implement a distributed dynamic scheduler of DAGs.
542 Distributed is somehow "contagious". If you start making distributed
543 decisions, there is no way to handle DAGs directly anymore (unless I
544 am missing something). You have to encode your DAGs in term of
545 communicating process to make the whole scheduling process
546 distributed. Here is an example of how you could do that. Assume T1
547 has to be done before T2.
550 int your_agent(int argc, char *argv[] {
552 T1 = MSG_task_create(...);
553 T2 = MSG_task_create(...);
557 if(cond) MSG_task_execute(T1);
559 if((MSG_task_get_remaining_computation(T1)=0.0) && (you_re_in_a_good_mood))
562 /* do something else */
568 If you decide that the distributed part is not that much important and that
569 DAG is really the level of abstraction you want to work with, then you should
570 give a try to \ref SD_API.
572 \section faq_dynamic Dynamic resources and platform building
574 \subsection faq_platform Building a realistic platform
576 We can speak more than an hour on this subject and we still do not have
577 the right answer, just some ideas. You can read the following
578 <a href="http://graal.ens-lyon.fr/~alegrand/articles/Simgrid-Introduction.pdf">slides</a>.
579 It may give you some hints. You can also have a look at the
580 <tt>tools/platform_generation/</tt> directory. There is a perl-script
581 we use to annotate a Tiers generated platform.
583 \subsection faq_SURF_dynamic How can I have variable resource availability?
585 A nice feature of SimGrid is that it enables you to seamlessly have
586 resources whose availability change over time. When you build a
587 platform, you generally declare CPUs like that:
590 <cpu name="Cpu A" power="100.00"/>
593 If you want the availability of "CPU A" to change over time, the only
594 thing you have to do is change this definition like that:
597 <cpu name="Cpu A" power="100.00" availability_file="trace_A.txt" state_file="trace_A_failure.txt"/>
600 For CPUs, availability files are expressed in fraction of available
601 power. Let's have a look at what "trace_A.txt" may look like:
610 At time 0, our CPU will deliver 100 Mflop/s. At time 11.0, it will
611 deliver only 50 Mflop/s until time 20.0 where it will will start
612 delivering 90 Mflop/s. Last at time 21.0 (20.0 plus the periodicity
613 1.0), we'll be back to the beginning and it will deliver 100Mflop/s.
615 Now let's look at the state file:
622 A negative value means "off" while a positive one means "on". At time
623 1.0, the CPU is on. At time 1.0, it is turned off and at time 2.0, it
624 is turned on again until time 12 (2.0 plus the periodicity 10.0). It
625 will be turned on again at time 13.0 until time 23.0, and so on.
627 Now, let's look how the same kind of thing can be done for network
628 links. A usual declaration looks like:
631 <network_link name="LinkA" bandwidth="10.0" latency="0.2"/>
634 You have at your disposal the following options: bandwidth_file,
635 latency_file and state_file. The only difference with CPUs is that
636 bandwidth_file and latency_file do not express fraction of available
637 power but are expressed directly in Mb/s and seconds.
639 \subsection faq_flexml_bypassing How can I have some C functions do what the platform file does?
641 So you want to bypass the XML files parser, uh? Maybe doin some parameter
642 sweep experiments on your simulations or so? This is possible, but it's not
643 really easy. Here is how it goes.
645 For this, you have to first remember that the XML parsing in SimGrid is done
646 using a tool called FleXML. Given a DTD, this gives a flex-based parser. If
647 you want to bypass the parser, you need to provide some code mimicking what
648 it does and replacing it in its interactions with the SURF code. So, let's
649 have a look at these interactions.
651 FleXML parser are close to classical SAX parsers. It means that a
652 well-formed SimGrid platform XML file might result in the following
655 - start "platform_description"
656 - start "cpu" with attributes name="host1" power="1.0"
658 - start "cpu" with attributes name="host2" power="2.0"
660 - start "network_link" with ...
662 - start "route" with ...
664 - start "route" with ...
666 - end "platform_description"
668 The communication from the parser to the SURF code uses two means:
669 Attributes get copied into some global variables, and a surf-provided
670 function gets called by the parser for each event. For example, the event
671 - start "cpu" with attributes name="host1" power="1.0"
673 let the parser do the equivalent of:
675 strcpy("host1",A_cpu_name);
680 In SURF, we attach callbacks to the different events by initializing the
681 pointer functions to some the right surf functions. Example in
682 workstation_KCCFLN05.c (surf_parse_open() ends up calling surf_parse()):
684 // Building the routes
685 surf_parse_reset_parser();
686 STag_route_fun=parse_route_set_endpoints;
687 ETag_route_element_fun=parse_route_elem;
688 ETag_route_fun=parse_route_set_route;
689 surf_parse_open(file);
690 xbt_assert1((!surf_parse()),"Parse error in %s",file);
694 So, to bypass the FleXML parser, you need to write your own version of the
695 surf_parse function, which should do the following:
696 - Call the corresponding STag_<tag>_fun function to simulate tag start
697 - Fill the A_<tag>_<attribute> variables with the wanted values
698 - Call the corresponding ETag_<tag>_fun function to simulate tag end
699 - (do the same for the next set of values, and loop)
701 Then, tell SimGrid that you want to use your own "parser" instead of the stock one:
703 surf_parse = surf_parse_bypass;
704 MSG_create_environment(NULL);
707 An example of this trick is distributed in the file examples/msg/msg_test_surfxml_bypassed.c
709 \section faq_limits Pushing the limits
711 \subsection faq_context_1000 I want thousands of simulated processes
713 SimGrid can use either pthreads library or the UNIX98 contextes. On most
714 systems, the number of pthreads is limited and then your simulation may be
715 limited for a stupid reason. This is especially true with the current linux
716 pthreads, and I cannot get more than 2000 simulated processes with pthreads
717 on my box. The UNIX98 contexts allow me to raise the limit to 25,000
718 simulated processes on my laptop.
720 The <tt>--with-context</tt> option of the <tt>./configure</tt> script allows
721 you to choose between UNIX98 contextes (<tt>--with-context=ucontext</tt>)
722 and the pthread version ( (<tt>--with-context=pthread</tt>). The default
723 value is ucontext when the script detect a working UNIX98 context
724 implementation. On Windows boxes, the provided value is discarded and an
725 adapted version is picked up.
727 We experienced some issues with contextes on some rare systems (solaris 8
728 and lower or old alpha linuxes comes to mind). The main problem is that the
729 configure script detect the contextes as being functional when it's not
730 true. If you happen to use such a system, switch manually to the pthread
731 version, and provide us with a good patch for the configure script so that
732 it is done automatically ;)
734 \subsection faq_context_10000 I want hundred thousands of simulated processes
736 As explained above, SimGrid can use UNIX98 contextes to represent and handle
737 the simulated processes. Thanks to this, the main limitation to the number
738 of simulated processes becomes the available memory.
740 Here are some tricks I had to use in order to run a token ring between
741 25,000 processes on my laptop (1Gb memory, 1.5Gb swap).
743 - First of all, make sure your code runs for a few hundreds processes
744 before trying to push the limit. Make sure it's valgrind-clean, ie that
745 valgrind does not report neither memory error nor memory leaks. Indeed,
746 numerous simulated processes result in *fat* simulation hindering debugging.
748 - It was really boring to write 25,000 entries in the deployment file, so I wrote
749 a little script <tt>examples/gras/tokenS/make_deployment.pl</tt>, which you may
750 want to adapt to your case. You could also think about hijacking
751 the SURFXML parser (have look at \ref faq_flexml_bypassing).
753 - The deployment file became quite big, so I had to do what is in the FAQ
754 entry \ref faq_flexml_limit
756 - Each UNIX98 context has its own stack entry. As debugging this is quite
757 hairly, the default value is a bit overestimated so that user don't get
758 into trouble about this. You want to tune this size to increse the number
759 of processes. This is the <tt>STACK_SIZE</tt> define in
760 <tt>src/xbt/context_private.h</tt>, which is 128kb by default.
761 Reduce this as much as you can, but be warned that if this value is too
762 low, you'll get a segfault. The token ring example, which is quite simple,
763 runs with 40kb stacks.
765 \section faq_troubleshooting Troubleshooting
767 \subsection faq_compil_trouble ./configure fails!
769 We now only one reason for the configure to fail:
771 - <b>You are using a borken build environment</b>\n
772 If symptom is that configure complains about gcc not being able to build
773 executables, you are probably missing the libc6-dev package. Damn Ubuntu.
775 If you experience other kind of issue, please get in touch with us. We are
776 always interested in improving our portability to new systems.
778 \subsection faq_distcheck_fails Dude! "make check" fails on my machine!
780 Don't assume we never run this target, because we do. Really. Promise!
782 There is several reasons which may cause the make check to fail on your
785 - <b>You are using a borken libc (probably concerning the contextes)</b>.\n
786 The symptom is that the "make check" fails within the examples/msg directory.\n
787 By default, SimGrid uses something called ucontexts. This is part of the
788 libc, but it's quite undertested. For example, some (old) versions of the
789 glibc on alpha do not implement these functions, but provide the stubs
790 (which return ENOSYS: not implemented). It fools our detection mecanism
791 and leads to segfaults.\n
792 On some x86_64, the pointer to function is stored into a integer, but int
793 are 32bits only on this arch while pointers are 64bits. Our detection
794 mecanism also fails to detect the problem, which leads to segfaults.\n
795 In both cases, there is not much we can do to fix the bug. We are working
796 on a workaround for x86_64 machines, but in the meanwhile, you can
797 compile with --with-context=pthread to avoid ucontext completely. You'll
798 be a bit more limitated in the number of simulated processes you can start
799 concurently, but 5000 processes is still enough for most purposes, isn't
801 This limitation is the reason why we insist on using this piece of ...
802 software even if it's so troublesome.\n
803 <b>=> use --with-pthread on AMD64 architecture that do not have an
804 ultra-recent libc.</b>
806 - <b>There is a bug in SimGrid we aren't aware of</b>.\n
807 If none of the above apply, please drop us a mail on the mailing list so
808 that we can check it out.
810 \subsection faq_longjmp longjmp madness in valgrind
812 This is when valgrind starts complaining about longjmp things, just like:
814 \verbatim ==21434== Conditional jump or move depends on uninitialised value(s)
815 ==21434== at 0x420DBE5: longjmp (longjmp.c:33)
817 ==21434== Use of uninitialised value of size 4
818 ==21434== at 0x420DC3A: __longjmp (__longjmp.S:48)
821 or even when it reports scary things like:
823 \verbatim ==24023== Warning: client switching stacks? SP change: 0xBE3FF618 --> 0xBE7FF710
824 x86->IR: unhandled instruction bytes: 0xF4 0xC7 0x83 0xD0
825 ==24023== to suppress, use: --max-stackframe=4194552 or greater
826 ==24023== Your program just tried to execute an instruction that Valgrind
827 ==24023== did not recognise. There are two possible reasons for this.
828 ==24023== 1. Your program has a bug and erroneously jumped to a non-code
829 ==24023== location. If you are running Memcheck and you just saw a
830 ==24023== warning about a bad jump, it's probably your program's fault.
831 ==24023== 2. The instruction is legitimate but Valgrind doesn't handle it,
832 ==24023== i.e. it's Valgrind's fault. If you think this is the case or
833 ==24023== you are not sure, please let us know.
834 ==24023== Either way, Valgrind will now raise a SIGILL signal which will
835 ==24023== probably kill your program.
837 ==24023== Process terminating with default action of signal 4 (SIGILL)
838 ==24023== Illegal opcode at address 0x420D234
839 ==24023== at 0x420D234: abort (abort.c:124)
842 This is the sign that you didn't used the exception mecanism well. Most
843 probably, you have a <tt>return;</tt> somewhere within a <tt>TRY{}</tt>
844 block. This is <b>evil</b>, and you must not do this. Did you read the section
847 \subsection faq_valgrind Valgrind spits tons of errors!
849 It may happen that valgrind, the memory debugger beloved by any decent C
850 programmer, spits tons of warnings like the following :
851 \verbatim ==8414== Conditional jump or move depends on uninitialised value(s)
852 ==8414== at 0x400882D: (within /lib/ld-2.3.6.so)
853 ==8414== by 0x414EDE9: (within /lib/tls/i686/cmov/libc-2.3.6.so)
854 ==8414== by 0x400B105: (within /lib/ld-2.3.6.so)
855 ==8414== by 0x414F937: _dl_open (in /lib/tls/i686/cmov/libc-2.3.6.so)
856 ==8414== by 0x4150F4C: (within /lib/tls/i686/cmov/libc-2.3.6.so)
857 ==8414== by 0x400B105: (within /lib/ld-2.3.6.so)
858 ==8414== by 0x415102D: __libc_dlopen_mode (in /lib/tls/i686/cmov/libc-2.3.6.so)
859 ==8414== by 0x412D6B9: backtrace (in /lib/tls/i686/cmov/libc-2.3.6.so)
860 ==8414== by 0x8076446: xbt_dictelm_get_ext (dict_elm.c:714)
861 ==8414== by 0x80764C1: xbt_dictelm_get (dict_elm.c:732)
862 ==8414== by 0x8079010: xbt_cfg_register (config.c:208)
863 ==8414== by 0x806821B: MSG_config (msg_config.c:42)
866 This problem is somewhere in the libc when using the backtraces and there is
867 very few things we can do ourselves to fix it. Instead, here is how to tell
868 valgrind to ignore the error. Add the following to your ~/.valgrind.supp (or
869 create this file on need). Make sure to change the obj line according to
870 your personnal mileage (change 2.3.6 to the actual version you are using,
871 which you can retrieve with a simple "ls /lib/ld*.so").
874 name: Backtrace madness
881 fun:__libc_dlopen_mode
884 Then, you have to specify valgrind to use this suppression file by passing
885 the <tt>--suppressions=$HOME/.valgrind.supp</tt> option on the command line.
886 You can also add the following to your ~/.bashrc so that it gets passed
887 automatically. Actually, it passes a bit more options to valgrind, and this
888 happen to be my personnal settings. Check the valgrind documentation for
891 \verbatim export VALGRIND_OPTS="--leak-check=yes --leak-resolution=high --num-callers=40 --tool=memcheck --suppressions=$HOME/.valgrind.supp" \endverbatim
893 \subsection faq_flexml_limit I get the message "surf_parse_lex: Assertion `next<limit' failed."
895 This is because your platform file is too big for the parser.
897 Actually, the message comes directly from FleXML, the technology on top of
898 which the parser is built. FleXML has the bad idea of fetching the whole
899 document in memory before parsing it. And moreover, the memory buffer size
900 must be determinded at compilation time.
902 We use a value which seems big enough for our need without bloating the
903 simulators footprints. But of course your mileage may vary. In this case,
904 just edit src/surf/surfxml.l modify the definition of
905 FLEXML_BUFFERSTACKSIZE. E.g.
908 #define FLEXML_BUFFERSTACKSIZE 1000000000
911 Then recompile and everything should be fine, provided that your version of
912 Flex is recent enough (>= 2.5.31). If not the compilation process should
915 A while ago, we worked on FleXML to reduce a bit its memory consumtion, but
916 these issues remain. There is two things we should do:
918 - use a dynamic buffer instead of a static one so that the only limit
919 becomes your memory, not a stupid constant fixed at compilation time
920 (maybe not so difficult).
921 - change the parser so that it does not need to get the whole file in
922 memory before parsing
923 (seems quite difficult, but I'm a complete newbe wrt flex stuff).
925 These are changes to FleXML itself, not SimGrid. But since we kinda hijacked
926 the development of FleXML, I can grant you that any patches would be really
927 welcome and quickly integrated.
929 \subsection faq_gras_transport GRAS spits networking error messages
931 Gras, on real platforms, naturally use regular sockets to communicate. They
932 are deeply hiden in the gras abstraction, but when things go wrong, you may
933 get some weird error messages. Here are some example, with the probable
936 - <b>Transport endpoint is not connected</b>: several processes try to open
937 a server socket on the same port number of the same machine. This is
938 naturally bad and each process should pick its own port number for this.\n
939 Maybe, you just have some processes remaining from a previous experiment
941 Killing them may help, but again if you kill -KILL them, you'll have to
942 wait for a while: they didn't close there sockets properly and the system
943 needs a while to notice that this port is free again.
945 - <b>Socket closed by remote side</b>: if the remote process is not
946 supposed to close the socket at this point, it may be dead.
948 - <b>Connection reset by peer</b>: I found this on internet about this
949 error. I think it's what's happening here, too:\n
950 <i>This basically means that a network error occurred while the client was
951 receiving data from the server. But what is really happening is that the
952 server actually accepts the connection, processes the request, and sends
953 a reply to the client. However, when the server closes the socket, the
954 client believes that the connection has been terminated abnormally
955 because the socket implementation sends a TCP reset segment telling the
956 client to throw away the data and report an error.\n
957 Sometimes, this problem is caused by not properly closing the
958 input/output streams and the socket connection. Make sure you close the
959 input/output streams and socket connection properly. If everything is
960 closed properly, however, and the problem persists, you can work around
961 it by adding a one-second sleep before closing the streams and the
962 socket. This technique, however, is not reliable and may not work on all
964 Since GRAS sockets are closed properly (repeat after me: there is no bug
965 in GRAS), it is either that you are closing your sockets on server side
966 before the client get a chance to read them (use gras_os_sleep() to delay
967 the server), or the server died awfully before the client got the data.
969 \subsection faq_deadlock There is a deadlock !!!
971 Unfortunately, we cannot debug every code written in SimGrid. We
972 furthermore believe that the framework provides ways enough
973 information to debug such informations yourself. If the textual output
974 is not enough, Make sure to check the \ref faq_visualization FAQ entry to see
975 how to get a graphical one.
977 Now, if you come up with a really simple example that deadlocks and
978 you're absolutely convinced that it should not, you can ask on the
979 list. Just be aware that you'll be severely punished if the mistake is
980 on your side... We have plenty of FAQ entries to redact and new
981 features to implement for the impenitents! ;)
983 \subsection faq_big_fat_warning A BIG FAT WARNING is reported telling me that my platform and deployment files are too old.
985 We have decided to change the units in SimGrid. Now we use Bytes, Flops and
986 seconds instead of MBytes, MFlops and seconds... Units should be updated
987 accordingly and the version of platform_description should be set to a
988 valuer greater than 1:
990 <platform_description version="1">
992 You should try to use the surfxml_update.pl script that can be found
993 <a href="http://gforge.inria.fr/plugins/scmcvs/cvsweb.php/contrib/platform_generation/?cvsroot=cvsroot%2Fsimgrid">here</a>.
995 \subsection faq_bugrepport So I've found a bug in SimGrid. How to repport it?
997 We do our best to make sure to hammer away any bugs of SimGrid, but this is
998 still an academic project so please be patient if/when you find bugs in it.
999 If you do, the best solution is to drop an email either on the simgrid-user
1000 or the simgrid-devel mailing list and explain us about the issue. You can
1001 also decide to open a formal bug report using the
1002 <a href="https://gforge.inria.fr/tracker/?atid=165&group_id=12&func=browse">relevant
1003 interface</a>. You need to login on the server to get the ability to submit
1006 We will do our best to solve any problem repported, but you need to help us
1007 finding the issue. Just telling "it segfault" isn't enough. Telling "It
1008 segfaults when running the attached simulator" doesn't really help either.
1009 You may find the following article interesting to see how to repport
1010 informative bug repports:
1011 http://www.chiark.greenend.org.uk/~sgtatham/bugs.html (it is not SimGrid
1012 specific at all, but it's full of good advices).
1014 \author Arnaud Legrand (arnaud.legrand::imag.fr)
1015 \author Martin Quinson (martin.quinson::loria.fr)