mirror of https://github.com/BOINC/boinc.git
37 lines
1.1 KiB
HTML
37 lines
1.1 KiB
HTML
|
<h2>Back end examples</h2>
|
||
|
|
|
||
|
|
<p>
|
||
|
|
Here are some examples of the "back ends" that
|
||
|
|
could be used by various types of project.
|
||
|
|
|
||
|
|
<p>
|
||
|
|
<b>SETI@home with local data servers</b>
|
||
|
|
<ul>
|
||
|
|
<li>
|
||
|
|
A <b>splitter program</b> generates data files.
|
||
|
|
For each data file, calls the BOINC server library to
|
||
|
|
create a workunit record and several result records,
|
||
|
|
and stores the data file on disk of a data server.
|
||
|
|
<li>
|
||
|
|
A <b>garbage collector</b> program deletes data files for which enough
|
||
|
|
results have been returned.
|
||
|
|
Generates new result records as needed.
|
||
|
|
<li>
|
||
|
|
A <b>result handler</b> program checks for new results.
|
||
|
|
Parses result files, and transfers contents to
|
||
|
|
a separate "science database".
|
||
|
|
Does redundancy checking, and marks result records as valid or invalid.
|
||
|
|
</ul>
|
||
|
|
|
||
|
|
<p>
|
||
|
|
<b>SETI@home with remote data servers</b>
|
||
|
|
<p>
|
||
|
|
As above, but the splitter program writes data files to DLT tape.
|
||
|
|
The result records are generated, using the
|
||
|
|
URL of a selected data server.
|
||
|
|
The tape is mailed to the data server.
|
||
|
|
An operator at the data server transfers the files to disk,
|
||
|
|
than uses a web control panel to mark the batch as ready,
|
||
|
|
which marks the result records as ready.
|
||
|
|
<p>
|