Merge branch 'master' into energy-pstate

[simgrid.git] / doc / doxygen / tutorial_msg.doc

/*! @page tutorial_msg SimGrid Tutorial with MSG

SimGrid is a toolkit providing the core functionalities for the
simulation of distributed applications in heterogeneous distributed
environments.

The project goal is both to facilitate research and to help improving
real applications in the area of distributed and parallel systems,
ranging from simple network of workstations to Computational Grids to
Clouds and to supercomputers.

\tableofcontents

\section  Scenario
The goal of this practical session is to illustrate various usage of
the MSG interface. To this end we will use the following simple setting:

> Assume we have a (possibly large) bunch of (possibly large) data to
> process and which originally reside on a server (a.k.a. master). For
> sake of simplicity, we assume all input file require the same amount
> of computation. We assume the server can be helped by a (possibly
> large) set of worker machines. What is the best way to organize the
> computations ?

Although this looks like a very simple setting it raises several
interesting questions:

- Which algorithm should the master use to send workload?

    The most obvious algorithm would be to send tasks to workers in a
    round-robin fashion. This is the initial code we provide you.

    A less obvious but probably more efficient approach would be to set up
    a request mechanism where a client first ask for tasks, which allows
    the server to decide which request to answer and possibly to send
    the tasks to the fastest machines. Maybe you can think of a
    smarter mechanism...

- How many tasks should the client ask for?

    Indeed, if we set up a request mechanism so that workers only
    send request whenever they have no more task to process, they are
    likely to be poorly exploited since they will have to wait for the
    master to consider their request and for the input data to be
    transferred. A client should thus probably request a pool of tasks
    but if it requests too many tasks, it is likely to lead to a poor
    load-balancing...

- How is the quality of such algorithm dependent on the platform
    characteristics and on the task characteristics?

    Whenever the input communication time is very small compared to
    processing time and workers are homogeneous, it is likely that the
    round-robin algorithm performs very well. Would it still hold true
    when transfer time is not negligible and the platform is, say,
    a volunteer computing system ?

- The network topology interconnecting the master and the workers
  may be quite complicated. How does such a topology impact the
  previous result?

    When data transfers are the bottleneck, it is likely that a good
    modeling of the platform becomes essential. In this case, you may
    want to be able to account for complex platform topologies.

- Do the algorithms depend on a perfect knowledge of this
  topology?

    Should we still use a flat master worker deployment or should we
    use a

- How is such an algorithm sensitive to external workload variation?

    What if bandwidth, latency and power can vary with no warning?
    Shouldn't you study whether your algorithm is sensitive to such
    load variations?

- Although an algorithm may be more efficient than another, how
  does it interfere with other applications?

    As you can see, this very simple setting may need to evolve way
    beyond what you initially imagined.

    <blockquote> Premature optimization is  the root of all evil. -- D.E.Knuth</blockquote>

    Furthermore, writing your own simulator is much harder than you
    may imagine. This is why you should rely on an established and flexible
    one.

The following figure is a screenshot of [triva][fn:1] visualizing a [SimGrid
simulation][fn:2] of two master worker applications (one in light gray and
the other in dark gray) running in concurrence and showing resource
usage over a long period of time.

![Test](./sc3-description.png)

\section Prerequisites

Of course, you need to install SimGrid before taking this tutorial.
Please refer to the relevant Section: \ref install.

## Tutorials

A lot of information on how to install and use Simgrid are
provided by the [online documentation][fn:4] and by several tutorials:

- http://simgrid.gforge.inria.fr/tutorials/simgrid-use-101.pdf
- http://simgrid.gforge.inria.fr/tutorials/simgrid-tracing-101.pdf
- http://simgrid.gforge.inria.fr/tutorials/simgrid-platf-101.pdf

## Installing the visualization softwares

Several tools can be used to visualize the result of SimGrid
simulations and get a better understanding of simulations.

- [pajeng][fn:5] provides a Gantt-chart visualization.
- [Vite][fn:6] also provides a Gantt-chart visualization.

Under Debian or Ubuntu, this is really easy with apt-get, while you
may have to install from the source on other systems. Check the
documentation of each software for more details.

~~~~{.sh}
sudo apt-get install pajeng vite
~~~~

\section intro_start Let's get started

\anchor intro_setup
## Setting up and Compiling

The corresponding source files can be obtained
[online on GitLab](https://gitlab.inria.fr/simgrid/simgrid/tree/master/doc/msg-tuto-src). 
If you find the right button on the top right of the interface, you can download the whole 
directory in one archive file. If you wish, you can find other platform file in 
[this GitLab directory](https://gitlab.inria.fr/simgrid/simgrid/tree/master/examples/platforms).

As you can see, there is already a little Makefile that compiles
everything for you. If you struggle with the compilation, then you should double check 
your @ref install "SimGrid installation". 
On need, please refer to the @ref install_yours_trouble section.

Once the tiny example has been compiled and it can be easily run as follows:

~~~~{.sh}
./masterworker0 platforms/platform.xml deployment0.xml
~~~~

For a more "fancy" output, you can use simgrid-colorizer. 

~~~~{.sh}
./masterworker0 platforms/platform.xml deployment0.xml 2>&1 | simgrid-colorizer
~~~~

If you installed SimGrid to a non-standard path, you may have to
specify the full path to simgrid-colorizer on the above line, such as
\c /opt/simgrid/bin/simgrid-colorizer. If you did not install it at all,
you can find it in <simgrid_root_directory>/bin/colorize.

For a classical Gantt-Chart visualization, you can produce a [Paje][fn:5] trace:

~~~~{.sh}
./masterworker0 platforms/platform.xml deployment0.xml --cfg=tracing:yes \
    --cfg=tracing/msg/process:yes
pajeng simgrid.trace
~~~~

Alternatively, you can use [vite][fn:6].

~~~~{.sh}
./masterworker0 platforms/platform.xml deployment0.xml --cfg=tracing:yes \
    --cfg=tracing/msg/process:yes --cfg=tracing/basic:yes
vite simgrid.trace
~~~~

## Getting Rid of Workers in the Deployment File

In the previous example, the deployment file `deployment0.xml`
is tightly connected to the platform file `platform.xml` and a
worker process is launched on each host:

~~~~{.xml}
<?xml version='1.0'?>
<!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
<platform version="3">
  <!-- The master process (with some arguments) -->
  <process host="Tremblay" function="master">
     <argument value="20"/>       <!-- Number of tasks -->
     <argument value="50000000"/>  <!-- Computation size of tasks -->
     <argument value="1000000"/>   <!-- Communication size of tasks -->
     <argument value="Jupiter"/>  <!-- First worker -->
     <argument value="Fafard"/>   <!-- Second worker -->
     <argument value="Ginette"/>  <!-- Third worker -->
     <argument value="Bourassa"/> <!-- Last worker -->
     <argument value="Tremblay"/> <!-- Me! I can work too! -->
  </process>
  <!-- The worker process (with no argument) -->
  <process host="Tremblay" function="worker" on_failure="RESTART"/>
  <process host="Jupiter" function="worker" on_failure="RESTART"/>
  <process host="Fafard" function="worker" on_failure="RESTART"/>
  <process host="Ginette" function="worker" on_failure="RESTART"/>
  <process host="Bourassa" function="worker" on_failure="RESTART"/>
</platform>
~~~~

This is ok as the platform is rather small but will be painful when
using larger platforms. Instead, modify the simulator
`masterworker0.c` into `masterworker1.c` so that the master
launches a worker process on all the other machines at startup. The
new deployment file `deployment1.xml` should thus now simply be:

~~~~{.xml}
<?xml version='1.0'?>
<!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
<platform version="3">
  <!-- The master process (with some arguments) -->
  <process host="Tremblay" function="master">
     <argument value="20"/>       <!-- Number of tasks -->
     <argument value="50000000"/>  <!-- Computation size of tasks -->
     <argument value="1000000"/>   <!-- Communication size of tasks -->
  </process>
</platform>
~~~~

To this end you may need the following MSG functions (click on the links
to see their descriptions):

~~~~{.c}
int MSG_get_host_number(void);
xbt_dynar_t MSG_hosts_as_dynar(void);
void * xbt_dynar_to_array (xbt_dynar_t dynar);
msg_process_t MSG_process_create(const char *name, xbt_main_func_t code,
                                 void *data, msg_host_t host);
~~~~

\note
    It may avoid bugs later to avoid launching a worker on
    the master host so you probably want to remove it from the host
    list.

The `data` field of the @ref MSG_process_create can be used to pass
a channel name that will be private between master
and workers (e.g., `master_name:worker_name`). Adding the
`master_name` in the channel name will allow to easily have several
masters and a worker per master on each machine. To this end, you
may need to use the following functions:

~~~~{.c}
msg_host_t MSG_host_self(void);
const char * MSG_host_get_name(msg_host_t host);
msg_process_t MSG_process_self(void);
void * MSG_process_get_data(msg_process_t process);
~~~~

If you are not too familiar with string
manipulation in C, you may want to use the following functions
(see the C reference for details):

~~~~{.c}
char *strcpy(char *dest, const char *src);
char *strcat(char *dest, const char *src);
~~~~

## Setting up a Time Limit Mechanism

In the current version, the number of tasks is defined through the
worker arguments. Hence, tasks are created at the very beginning of
the simulation. Instead, create tasks as needed and provide a time
limit indicating when it stops sending tasks. To this end, you will
obviously need to know what time it is:

~~~~{.c}
double MSG_get_clock(void);
~~~~

Otherwise, a quite effective way of terminating the simulation
would be to use some of the following functions:

~~~~{.c}
void MSG_process_kill(msg_process_t process);
int MSG_process_killall(int reset_PIDs);
~~~~

Anyway, the new deployment `deployment2.xml` file should thus look
like this:

~~~~{.xml}
<?xml version='1.0'?>
<!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
<platform version="3">
  <process host="Tremblay" function="master">
     <argument value="3600"/>      <!-- Simulation timeout -->
     <argument value="50000000"/>  <!-- Computation size of tasks -->
     <argument value="1000000"/>   <!-- Communication size of tasks -->
  </process>
</platform>
~~~~

It may also be a good idea to transform most of the `XBT_INFO` into
`XBT_DEBUG` (e.g., keep the information on the total number of
tasks processed). These debug messages can be activated as follows:

~~~~{.sh}
./masterworker2 platforms/platform.xml deployment2.xml --log=msg_test.thres:debug
~~~~

## Using the Tracing Mechanism

SimGrid can trace all resource consumption and the outcome can be
displayed as illustrated in the section \ref intro_setup. However, when several
masters are deployed, it is hard to understand what happens.

~~~~{.xml}
<?xml version='1.0'?>
<!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
<platform version="3">
  <process host="Tremblay" function="master">
     <argument value="3600"/>      <!-- Simulation timeout -->
     <argument value="50000000"/>  <!-- Computation size of tasks -->
     <argument value="10"/>   <!-- Communication size of tasks -->
  </process>
  <process host="Fafard" function="master">
     <argument value="3600"/>      <!-- Simulation timeout -->
     <argument value="50000000"/>  <!-- Computation size of tasks -->
     <argument value="10"/>   <!-- Communication size of tasks -->
  </process>
  <process host="Jupiter" function="master">
     <argument value="3600"/>      <!-- Simulation timeout -->
     <argument value="50000000"/>  <!-- Computation size of tasks -->
     <argument value="10"/>   <!-- Communication size of tasks -->
  </process>
</platform>
~~~~

So let's use categories to track more precisely who does what and when:

~~~~{.c}
void TRACE_category(const char *category);
void MSG_task_set_category (msg_task_t task, const char *category);
~~~~

The outcome can then be visualized as a Gantt-chart as follows:

~~~~{.sh}
./masterworker3 platforms/platform.xml deployment3.xml --cfg=tracing:yes \
    --cfg=tracing/msg/process:yes
pajeng simgrid.trace
~~~~

Right now, you should realize that nothing is behaving like you expect. Most
workers are idle even though input data are ridiculous and there are several
masters deployed on the platform. So it should now be obvious that round robin
is actually very bad.

## Improving the Scheduling

Instead of a round-robin scheduling, let's implement a first-come
first-served mechanism. To this end, workers need to send a tiny
request first. A possible way to implement such a request with MSG
is to send on a specific channel (e.g., the name of the master
name) a task with payload 0 and whose attached data is the worker
name. This way, the master can keep track of which workers are idle
and willing to work.

To know whether it has pending requests, the master can use the
following [function][fn:7]:

~~~~{.c}
int MSG_task_listen(const char *alias);
~~~~

If so, it should get the request and push the corresponding host
into a dynar so that they can later be retrieved when sending a
real [task][fn:7].

~~~~{.c}
xbt_dynar_t xbt_dynar_new(const unsigned long elm_size,
                          void_f_pvoid_t const free_f);
void xbt_dynar_push(xbt_dynar_t const dynar, const void *src);
void xbt_dynar_shift(xbt_dynar_t const dynar, void *const dst);
unsigned long xbt_dynar_length(const xbt_dynar_t dynar);
~~~~

As you will soon realize, with such simple mechanisms, simple
deadlocks will soon appear. They can easily be removed with a
simple polling mechanism, hence the need for the following
[function][fn:7]:

~~~~{.c}
msg_error_t MSG_process_sleep(double nb_sec);
~~~~

As you should quickly realize, on the simple previous example, it
will double the throughput of the platform but will be quite
ineffective when input size of the tasks is not negligible anymore.

From this, many things can easily be added. For example, you could:
- add a performance measurement mechanism;
- enable the master to make smart scheduling choices using
  measurement information;
- allow workers to have several pending requests so as to overlap
  communication and computations as much as possible;
- ...

## Using More Elaborate Platforms

SimGrid offers a rather powerful platform modeling mechanism. The
`src/examples/platforms/` repository comprises a variety of platforms ranging
from simple to elaborate. Associated to a good
visualization tool to ensure your simulation is meaningful, they
can allow you to study to which extent your algorithm scales...

What is the largest number of tasks requiring 50e6 flops and 1e5
bytes that you manage to distribute and process in one hour on
`g5k.xml` (you should use `deployment_general.xml`)?

\section intro_todo TODO: Points to improve for the next time

- Propose equivalent exercises and skeleton in java.
- Propose a virtualbox image with everything (simgrid, pajeng, ...) already set
  up.
- Ease the installation on mac OS X (binary installer) and
  windows.
- Explain that programming in C or java and having a working
  development environment is a prerequisite.

[fn:1]: http://triva.gforge.inria.fr/index.html
[fn:2]: http://hal.inria.fr/inria-00529569
[fn:3]: http://hal.inria.fr/hal-00738321
[fn:4]: http://simgrid.gforge.inria.fr/simgrid/latest/doc/
[fn:5]: https://github.com/schnorr/pajeng/
[fn:6]: http://vite.gforge.inria.fr/





*/