Archive Structure Archive Structure
When data is being brought into the archive, it is convenient to place it in a directory tree which is broken down on a per-site basis. This is termed the mirror tree. Each directory in this tree contains all the data that is copied from a remote site. The directory name is the site's remote address. For example, any data from the site net.tamu.edu would be stored in the directory ..../mirrors/net.tamu.edu. Depending on how the data is mirrored, this directory may itself contain subdirectories.
This layout of data is convenient for mirroring, but is difficult to navigate. A second directory tree is built which contains pointers into the mirror tree - this is the subject tree. This tree is structured by topic.
Why choose this layout?
This format was chosen to simplify the mirror process, yet still allow
the data that was mirrored to be accessed intelligently. The following
scenario is very common when a site is mirrored: certain useful files
are copied from the site which correspond to different tools or
documents. All these files are placed in the same directory in the
mirrors tree. But these tools may perform different functions, or the
documents may realte to different subjects. We create links to the
individual tools from the directories in the subject tree. This
allows, for example, all documents related to viruses to be linked
under the pub/doc/viruses directory.
Built by Mark
Crosbie and
Ivan Krsul.
Security Archive Homepage.
COAST Project Page.
Purdue CS Dept page.