Workunits are purged only when their input files have been deleted. Because of BOINC's file-deletion policy, this implies that all results are completed. So when a workunit is purged, all its results are purged too.
Run db_purge from the project's bin/ directory. It will create an archive/ directory and store archive files there.
db_purge is normally run as a daemon, specified in the config.xml file. It has the following command-line options: "; list_start(); list_item("-min_age_days N", "Purge only WUs with mod_time at least N days in the past. Recommended value: 7 or so. This lets users examine their recent results." ); list_item("-max N", "Purge at most N WUs, then exit"); list_item("-max_wu_per_file N", "Write at most N WUs to each archive file. Recommended value: 10,000 or so." ); list_item("-zip", "Compress archive files using zip"); list_item("-gzip", "Compress archive files using gzip"); list_item("-d N", "Set logging verbosity to N (1,2,3)"); list_end(); echo "
The archive files have names of the form wu_archive_TIME and result_archive_TIME where TIME is the Unix time the file was created. In addition, db_purge generates index files 'wu_index' and 'result_index' associating each WU and result ID with the timestamp of its archive file.
The format of both type of index files is a number of rows each containing:
ID TIMEThe ID field of the WU or result, 5 spaces, and the timestamp part of the archive filename where the record with that ID can be found.
The format of a record in the result archive file is:
".html_text("