[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Everywhere Network



A Forbes story on a project called OceanStore:

http://www.forbes.com/futuretech/forbes/2001/0108/242.html

"The idea, dubbed OceanStore, is an architecture for the next  next-generation Internet. It is aimed
at solving one of the Net's looming big problems: how to reliably and securely retrieve data from
anywhere in the world, on any computing device."

You can read more about the project at its own site:
http://oceanstore.cs.berkeley.edu/

I find the Forbes story to be quite overstated in its claims about OceanStore, but it points out an
important trend toward distributed storage.  There are a large number of projects and systems now that
are based on this idea, and they will become increasingly important to archivists who hope to assist
in the preservation of electronic materials.

I've listed some of the most prominent ones at:
http://www-personal.si.umich.edu/~calz/ermlinks/diststor.htm

A common label for these sorts of systems is peer-to-peer (P2P), but that term is so widely used that
it's often difficult to define its boundaries.  One attempt to define the term can be found at
http://www.oreillynet.com/pub/a/p2p/2000/11/24/shirky1-whatisp2p.html

Some categories I would include:

- P2P computing - This involves actually executing programs at a distance, and includes initiatives
such as SOAP and XML-RPC.  Another variant of this category are systems that allow individuals to
donate computing cycles on their own computers in order tackle huge mathematically intensive
problems.  Entire operating systems that are distributed, such as Plan 9 from Bell Labs, could also
fit here.
- Distributed storage but central index - Napster is the most popular version of this category.  One
or more copies of a given file are available on people's individual hard drives around the world, but
initially discovering the file must be carried out through a central entity.
- Distribute storage and index - Gnutella has received the most press in this category.  It's similar
to Napster, but there's no central index entity.  Even that component is distributed.  Older systems
like AFS, that support sharing of a directory across multiple locations, could also (sort of) fit into
this category, though they usually involve only one copy of a file in a given physical location.
- Systems focusing on guaranteed availability - In order to ensure that a given digital object will
persist over time, many initiatives are taking the approach of keeping copies available in multiple
locations (sort of like keeping several copies of microfilm in various locations).  Freenet is a
popular example.  The LOCKSS (Lots of Copies Keeps Stuff Safe) project has preservation of access as
its primary goal, which is quite unusual in this category.  Many systems, including FreeNet, are based
on the gradual disappearance of files from the system that haven't been used for a given period of
time.  Some approaches keep entire copies of an entire digital object in various locations, while
others keep small chunks of them in various locations, from which the entire object can be
reconstructed when necessary.


=======================================================
 Cal Lee
 University of Michigan
 School of Information
 Phone: 734-647-0505
 http://www-personal.si.umich.edu/~calz/
 "Those who forget the past are condemned to reload it."
                        - Nick Montfort, July 2000

A posting from the Archives & Archivists LISTSERV List!

To subscribe or unsubscribe, send e-mail to listserv@listserv.muohio.edu
      In body of message:  SUB ARCHIVES firstname lastname
                    *or*:  UNSUB ARCHIVES
To post a message, send e-mail to archives@listserv.muohio.edu

Or to do *anything* (and enjoy doing it!), use the web interface at
     http://listserv.muohio.edu/archives/archives.html

Problems?  Send e-mail to Robert F Schmidt <rschmidt@lib.muohio.edu>