On Thu, Sep 29, 2011 at 12:12 PM, Sanne Grinovero <sanne@infinispan.org> wrote:

That's something that must be done too, but I'd suggest to start
creating a test which you can easily run in two JVMs on the same
"laptop-class" node, but having "real world" data and usage patterns.
For example you could load the Wikipedia in Infinispan, map each page
to an indexed POJO and then run queries on it to find articles, that
would be a very nice demo too.
After you got that working, it's going to be easier to actually start
a dozen of nodes in a physical cluster - we can start such a test but
since the labs are shared with other projects time is limited and so
we need to come with a plan and with the code well tested, scripting
as much as possible to automate the preparation phase (like wikipedia
download and upload to Infinispan).

Would you like to start looking into this?