Big Grid Specification - Draft 0.1Big Grid Specification - Draft 0.1The Big Grid is an OpenSource project for the development of a multi-purpose grid computing architecture.
The scale of the project is limited only by the willing participants.
This will do exactly what it says on the tin: attempt to prove the basic concept of the project. It will also allow the project to make estimates into the capability of a running system.
Its functional intention is to discover the host client's capability and report this to the central system. This data will then be aggregated and the totals reported back to the client. It is intended that this data always be available to the client applications.
This phase is where the real guts will begin to be mapped out.
The release of the first code.
The release of the first multi-tier code (supernode concept).
Currently the intended architecture will implement a three-tier system. This comprises the GPU, a series of Supanodes, and then all the clients that actually do all the calculation.
The BigGrid management services will run on the GPU. This is where jobs are initially submitted and the distribution of all job segments starts. Once completed, this is where the results makes their way back via the supanodes.
In definition process.
In definition process.
In definition process.
The system should offer the ability to alert the user if there is a newer version of the client currently in use.
The system may offer the ability to automatically update itself if there are updates available. This shall always be optional.
Intention will be expressed by the following terms:
Copyright
©2004 Big Grid