Shame on trust in distributed systems

Abstract

Approaches for building secure, distributed systems have fundamental limitations that prevent the construction of dynamic, Internet-scale systems. In this paper, we propose a concept of a shared reference monitor or Shamon that we believe will provide a basis for overcoming these limitations. First, distributed systems lack a principled basis for trust in the trusted computing bases of member machines. In most distributed systems, a trusted computing base is assumed. However, the fear of compromise due to misconfiguration or vulnerable software limits the cases where this assumption can be applied in practice. Where such trust is not assumed, current solutions are not scalable to large systems [7, 20]. Second, current systems do not ensure the enforcement of the flexible, distributed system security goals. Mandatory access control (MAC) policies aim to describe enforceable security goals, but flexible MAC solutions, such as SELinux, do not even provide a scalable solution for a single machine (due to the complexity of UNIX systems), much less a distributed system. A significant change in approach is necessary to develop a principled trusted computing base that enforces system security goals and scales to large distributed systems.