[OpenAFS-devel] Regarding GSoC 2010 Collaborative Caching Project

Matt W. Benjamin matt@linuxbox.com
Sat, 17 Apr 2010 05:50:59 -0400 (EDT)


Hi,
----- "shruti jain" <shruti.jain1988@gmail.com> wrote:

> Here is what I know about the cache manager and its file server
> interactions.

A client knows it has the most recent version of an object by comparing its data version with that of a cached copy.  A changed version is the typical reason for fs to send a callback, but not the only one, it turns out.

> 
> In the second part of the project, we can allow discovery of peer
> clients for collaboration. This can be done by modifying the file
> server to keep access logs of the clients and if a client requests for
> any data then its corresponding clients in the logs would be returned
> to the requesting clients. 

It sounds like you're interested in a really decentralized model.  This might end up quite different from what I understand the Michigan intermediate file server (and the Hades thing Simon posted on last year) was doing.  It might be worth pointing out that some of the wins Steve was mentioning seem to be natural wins in the intermediate model, without a lot of adaptation.

Matt


-- 

Matt Benjamin

The Linux Box
206 South Fifth Ave. Suite 150
Ann Arbor, MI  48104

http://linuxbox.com

tel. 734-761-4689
fax. 734-769-8938
cel. 734-216-5309