Browse Prior Art Database

A method for scheduling load and replication of data in a compute cloud using geographical positioning information to avoid geographics related risks or benefit from geograhics related advantages .

IP.com Disclosure Number: IPCOM000200003D
Publication Date: 2010-Sep-23
Document File: 2 page(s) / 47K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method for scheduling load and replication of data in a compute cloud using geographical positioning information to avoid geographics related risks or benefit from geograhics related advantages.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 2

A method for scheduling load and replication of data in a compute cloud using geographical positioning information to avoid geographics related risks or benefit from geograhics related advantages .

Today, there is a trend towards geographically distributed compute clouds. Also, in the design of Data centers, it is important to maintain data and resources at different geographical locations, to avoid loss of data or transactions or the ability to serve loads, in case of a calamity. In our attempt to avoid this loss, it is a human task to design how load and data are distributed over various data centers or the compute cloud. It would be beneficial if the compute cloud could define itself where systems are located and based on that, how load and data should be distributed to avoid damage in case of a calamity.

    Known solutions are to count the number network switches between two systems, assuming that they are further apart if there are more switches in between. Then data is replicated between those (storage) systems that are furthest apart. The drawback of this is that geographical area's, even when apart, may suffer from the same geophysical or meteorological event, e.g. in the case of an earthquake when two area's are on the same fault in the earth, or when two locations are in the route of the same hurricane, or when they are in the same flooding area etc. No autonomous system today is aware of such relationships.

    In our invention we add the geographical location as a property of a resource in a cloud (server, storage, network switch etc.). Using this location and a GIS based information system, the cloud can itself define how to distribute load over its resources e.g. in a group of data centers.

Examples:

The database can contain information on fault in the earth, flood prone areas. The system can be connected to a weather forecasting system, so it can move data awayfrom an area where a hurricane/tsunami is expected.

This allows for the dynamic creation of bac...