boinc/doc/api.html

<title>The BOINC Application Program Interface (API)</title> 
<h2>The BOINC Application Programming Interface (API)</h2>

The BOINC API is a set of C++ functions.
Unless otherwise specified,
the functions return an integer error code; zero indicates success.
The graphics API is described <a href=graphics.html>separately</a>.

<h3>Initialization and termination</h3>
The application must call
<pre>
    int boinc_init();
</pre>
before calling other BOINC functions or doing I/O.
It may call
<pre>
    struct APP_INIT_DATA {
        char project_preferences[4096];
        char user_name[256];
        char team_name[256];
        double wu_cpu_time;           // cpu time from previous sessions
        double total_cobblestones;
        double recent_avg_cobblestones;
    };

    int boinc_get_init_data(APP_INIT_DATA&);
</pre>
to get the following information:
<ul>
<li> <b>project_preferences</b>: An XML string containing
the user's project-specific preferences.
<li> <b>user_name</b>: the user's "screen name" on this project.
<li> <b>team_name</b>: the user's team name, if any.
<li> <b>wu_cpu_time</b>: the CPU time spent on this WU so far
<li> <b>total_cobblestones</b>: the user's total work for this project.
<li> <b>recent_avg_cobblestones</b>: the recent average work per day.
</ul>
<p>
These items might be used by the application in its graphics.
At any time it may call
<pre>
    double boinc_cpu_time();
</pre>
to get its current CPU time.
<p>
When the application has completed it must call
<pre>
    int boinc_finish(int status);
</pre>
<tt>status</tt> is nonzero if an error was encountered.

<h3>Resolving file names</h3>
Applications that use named input or output files must call
<pre>
    int boinc_resolve_filename(char *logical_name, char *physical_name);</tt>
</pre>
to convert logical file names to physical names.
For example, instead of
<pre>
    f = fopen("my_file", "r");
</pre>
</p>
the application might use
<p>
<pre>
    char resolved_name[256];
    retval = boinc_resolve_filename("my_file", resolved_name);
    if (retval) fail("can't resolve filename");
    f = fopen(resolved_name, "r");
</pre>
<tt>boinc_resolve_filename()</tt> doesn't need to be used for temporary files.

<h3>Checkpointing</h3>

Computations that use a significant amount of time
per work unit may want to periodically write the current
state of the computation to disk.
This is known as <b>checkpointing</b>.
The state file must include everything required
to start the computation again at the same place it was checkpointed.
On startup, the application
reads the state file to determine where to begin computation.
If the BOINC client quits or exits,
the computation can be restarted from the most recent checkpoint.

<p>
Do use checkpoint, an application should write to output and
state files using the <tt>MFILE</tt> class.
<pre>
class MFILE {
public:
    int open(char* path, char* mode);
    int _putchar(char);
    int puts(char*);
    int printf(char* format, ...);
    size_t write(const void* buf, size_t size, size_t nitems);
    int close();
    int flush();
};
</pre>
MFILE buffers data in memory
and writes to disk only on <tt>flush()</tt> or <tt>close()</tt>.
This lets you write output files and state files
more or less atomically.
Frequency of checkpointing is a user preference
(e.g. laptop users might want to checkpoint infrequently).
An application must call
<pre>
    bool boinc_time_to_checkpoint();
</pre>
whenever it reaches a point where it is able to checkpoint.
If this returns true,
the application should write the state file and flush all output files,
then call
<pre>
    void boinc_checkpoint_completed();
</pre>
A call to <tt>boinc_time_to_checkpoint()</tt> is extremely fast,
so there is little penalty in calling it frequently.

<h3>Fraction done</h3>
The core client GUI displays the percent done of workunits in progress.
To keep this display current, an application should
periodically call
<pre>
   boinc_fraction_done(double fraction_done);
</pre>
The <tt>fraction_done</tt> argument is a rough estimate of the
workunit fraction complete (0 to 1).
This function is extremely fast and can be called often.

<h3>Multi-program applications</h3>
Some applications consist of multiple programs:
a <b>main program</b> that acts as coordinator,
and one or more subsidiary programs.
Each program should use the BOINC API as described above.
<p>
Each program should have its own state file;
the state file of the coordinator program records
which subsidiary program was last active.
<p>
To correctly implement fraction done,
the main program should pass information to subsidiary programs
(perhaps as command-line arguments) indicating the starting and ending
fractions for that program.
<p>
The coordinator must call
<pre>
    void boinc_child_start();
</pre>
prior to forking a child process.
When the child is done, the coordinator
must get the child's CPU time, then call
<pre>
    void boinc_child_done(double total_cpu);
</pre>
before forking the next child process.


<hr>
<h3>Implementation</h3>
<p>
Application are executed in separate "catbox" directories,
allowing them to create and use temporary files without name conflicts.
Input and output files are kept outside the catbox.
The mappings from virtual to physical filenames use
"symbolic link" files in the catbox directory.
The name of such a file is the virtual name,
and the file contains an XML tag with the physical name.
(This scheme is used because of the lack of filesystem links in Windows.)

<p>
Communication between the core client and applications
is done through XML files in the catbox directory.
Several files are used.
<p>
<b>Files created by the core client, read by the app:</b>
(Once, at start of app)
<ul>
<li> Symbolic link files
<li> <b>fd_init.xml</b>:
specifies the mappings of file descriptors (stdin/stdout/stderr)
to physical files.

<li> <b>init_data.xml</b>: this contains the initialization data
returned by <tt>boinc_init()</tt> (see above),
as well as the minimum checkpoint period.
</ul>
<p>
<b>Files created by the API implementation, read by the core client:</b>
<ul>
<li>
<b>fraction_done.xml</b>:
contains the WU fraction done and the current CPU time from start of WU.
Written by the timer routine as needed.

<li>
<b>checkpoint_cpu.xml</b>
CPU time (from start of WU) at last checkpoint.
Written by checkpoint_completed.

</ul>
<p>
The API implementation uses a timer (60Hz);
the real-time clock is not available to applications.
This timer is used for several purposes:
<ul>
<li> To tell the app when to checkpoint;
<li> To regenerate the fraction done file
<li> To refresh graphics
</ul>

<p>
<b>Exit status</b>
The core client does a wait() to get the status.
boinc_finish() ends with an exit(status);
<p>
<b>Accounting of CPU time</b>:
(note: in Unix, a parent can't get the CPU time of a child
until the child exits.  So we're forced to measure it in the child.)
The core passes the WU CPU time in init_data.xml.
boinc_checkpoint_done() and boinc_finish() compute the new WU CPU time,
and write it to checkpoint_cpu.xml.
The core deletes this after reading.
If on exit there is no checkpoint_cpu.xml, it means the app
called exit(0) rather than boinc_finish().
In this case the core measures the child CPU itself.
<p>
The core client maintains 
<p>
<b>Timing of checkpoints</b>
<p>
The app library maintains time_until_checkpoint,
decremented from the timer handler.
boinc_time_to_checkpoint() returns true if this is zero or less.
boinc_checkpoint_done() resets it.

<p>
<b>Maintaining fraction done and current CPU</b>
<p>
These two quantities are transferred from the app library to
the core client in the file fraction_done.xml.
The parameter <tt>time_until_fraction_done_update</tt>,
passed in the initialization file,
determines how often this file is written.
It is written from the timer handler.
<p>
For multi-program applications, only the active application
must write the file.
The functions boinc_child_start() and boinc_child_done()
tell the app library to stop and start writing the file.
<p>
TO DO: this creates disk traffic.
Either figure out a way of increasing the period for users who don't
want disk access, or don't use disk files.