This paper sketches the design of the Eliot File System (Eliot), a mutable file system that maintains the pure immutability of its peer-to-peer (P2P) substrate by isolating mutation in an auxiliary metadata service. The immutability of address-to-content bindings has several advantages in P2P systems. However mutable file systems are desirable because they allow clients to update existing files; a necessary property for many applications. In order to facilitate modifications, the file system must provide some atom of mutability. Since this atom of mutability is a fundamental characteristic of the file system and not the underlying storage substrate, it is a mistake to violate the integrity of the substrate with special cases for mutability. Instead, Eliot employs a separate, generalized metadata service that isolates all mutation and client state in an auxiliary replicated database. Eliot provides fine-granularity file updates with either AFS open-close or NFS-like consistency semantics. Eliot builds a mutable filesystem on a global resource bed of purely immutable P2P block storage.

At the 1999 Workshop on Hot Topics in Operating Systems (HotOS VII), the attendees reached consensus that the most important issue facing the OS research community was "No-Futz" computing; eliminating the ongoing "futzing" that characterizes most systems today. To date, little research has been accomplished in this area. Our goal in this paper is to focus the research community on the challenges we face if we are to design systems that are truly futz-free, or even low-futz.