[Pipet Devel] Re: distributed filesystem proposal

J.W. Bizzaro bizzaro at geoserve.net
Mon Apr 10 20:11:10 EDT 2000


"J.W. Bizzaro" wrote:
> 
> (3) VOLUMES AND TEMPORARY WORKSPACES

Volumes are defined as completely isolated networks.  IOW, they have no inputs
or outputs.  They are the highest level of networks.  They could be simply
called 'networks' if all the networks they contain are called 'subnets'.  I
think this is how Overflow defines these.

Also, a volume (or 'network') is an XML description of connections between
subnets and nodes, as we have been saying all along.  And this description
resides on ONE COMPUTER (or in one BL).

Brad and I were just talking about this over chat.  We think a connection made
to find out what volumes are available will be done via DL -> Internet -> DL
-> BL connection.  Mounting, opening and locking volumes are handled the same
way.

What differs from this mechanism is how network execution is handled.  Of
course, the source to a node (a program for example) can be at yet another
location.  In such a case, the server that stores the volume needs to make a
connection to the other server.  But I'm wondering if it is the remote server
or the local server that should be doing the execution of the network in the
volume.  BOTH have copies of the network, correct?

Jeff
-- 
                      +----------------------------------+
                      |           J.W. Bizzaro           |
                      |                                  |
                      | http://bioinformatics.org/~jeff/ |
                      |                                  |
                      |        BIOINFORMATICS.ORG        |
                      |           The Open Lab           |
                      |                                  |
                      |    http://bioinformatics.org/    |
                      +----------------------------------+




More information about the Pipet-Devel mailing list