1) it is possible to archive incomplete (corrupt) files into the repository. Files should be tested as valid tar archives prior to archiving. This will substantially increase the archive times, however since it happens just once it should be manageable.
2) Interrupted processing can lead to corrupted/incomplete files where the file exists but is either empty or not readable at all. Searching through the archive for such files is impractical, however there could be a test mechanism in the project function that will send the file to quarantine if it encounters a read problem.
Two related issues:
1) it is possible to archive incomplete (corrupt) files into the repository. Files should be tested as valid tar archives prior to archiving. This will substantially increase the archive times, however since it happens just once it should be manageable.
2) Interrupted processing can lead to corrupted/incomplete files where the file exists but is either empty or not readable at all. Searching through the archive for such files is impractical, however there could be a test mechanism in the project function that will send the file to quarantine if it encounters a read problem.