Accountability for distributed systems

ABSTRACT

Social expectations play an important role in distributed systems that span multiple administrative domains. For instance, participants in peer-to-peer systems are expected to contribute resources for the common good; members of federated systems are expected to adhere to best practices and fulfil contractual obligations; and providers of hosting services are expected to respect the confidentiality and integrity of customers' data and computation. In society, *accountability* is widely used to incentivize and reward good performance, to expose failures and unwanted behavior, and to build trust among competing individuals and organizations. In this talk, I'll suggest that accountability is also a powerful tool in the design of distributed systems. Accountability allows good nodes to prove their past compliance and ensures that (intended or unintended) deviations by any node from the expected behavior are detectable. Accountability complements fault tolerance techniques and offers an alternative to these techniques in systems that provide best-effort service.

I'll outline the requirements and challenges for an accountable distributed system. We look at a definition of accountability that is strong enough to be useful, yet allows an efficient and practical implementation. Finally, I'll sketch an implementation of accountability that is applicable to a large class of distributed systems and scales to large numbers of nodes. Joint work with Andreas Haeberlen, Petr Kuznetsov and Rodrigo Rodrigues.

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.2 COMPUTER-COMMUNICATION NETWORKS
      C.2.4 Distributed Systems
          Subjects: Distributed applications

Additional Classification:
  D. Software
  D.4 OPERATING SYSTEMS
      D.4.5 Reliability
          Subjects: Fault-tolerance

General Terms:
Algorithms, Design, Reliability

Keywords:
accountability, distributed systems