An Archive Maintenance Language

Page Content

Ivan Krsul

Nov 11, 1994

Abstract

The Security Archive maintained by the Computer Operations, Audit, and
Security Technologies (COAST) group at Purdue University consists of
several thousand files that must be closely monitored by the Security
Archive Maintainers. Emphasis must be placed on data integrity,
accountability, non-repudiation, packaging, package version control,
and the association of high level semantic information to files,
collections of files, packages, or collections of packages.

The initial tools written to automatically maintain the archive were
rather crude and would only detect inconsistencies in the archive
before it deteriorated beyond recovery. Part of the archive operation
was to mirror remote sites; information would be copied from other
archive sites in an organized manner and would be periodically updated
to ensure consistency with the remote site. Keeping track of
several thousand mirrored files that could randomly appear and
disappear made the management of this archive even more difficult.

In our efforts to keep the archive running smoothly, we determined
that our small set of tools (mostly shell and PERL scripts) were not
sufficiently powerful. We set out to design a new system by examining
the key aspects of the data we wished to manage. The ultimate product
would be a system that would allow archive maintainers and system
administrators to manage large volumes of information.

In the talk we will present a high level view of the problems we wish
to solve, and a tentative solution to the problems with the definition
of a high level language specifically designed for the purpose.

Footer

Intranet Site