From: Slavik Krassovsky (viatk_at_windows.microsoft.com)
Date: Wed Mar 03 2004 - 17:06:22 PST
This rather recent paper discusses a peer-to-peer system that is
designed to function as an Internet-wide persistent storage entity. This
entity is organized as a collection of nodes, each of which hosts files.
A routing layer called Pastry is discussed briefly.
A simple and clean interface is provided for insertion, retrieval and
reclamation of the file. This interface is based on a 160-bit values
called fileId - It's interesting to note that fileId essentially is a
capability to the file.
Overall I liked this paper a lot - although I find that reading
efficiency should be considered before accurate utilization of the
storage.
This archive was generated by hypermail 2.1.6 : Wed Mar 03 2004 - 17:06:32 PST