GRIDCLOUD PROJECT WEB SITE

This codeplex site is used to distribute source and binary codes for the Cornell/WSU GridCloud platform, which was created under ARPAe (GENI program) support.  The purpose of GridCloud is to collect data in real-time from the transmission network of a large-scale power grid, archive that data with the associated network model, and then support such tasks as real-time state estimation, contingency analysis, what-if scenario exploration, etc.  The project is expected to complete early in 2015, and this site will not be fully populated with system code and platform elements until fairly late in the development cycle. 

GridCloud is designed with security and high assurance properties such as real-time performance and scalability in mind.  The system would normally be used on a cloud platform such as Amazon EC2 but can easily be ported to other kinds of "virtual machine" environments.  With our help, it can be configured to co-exist with firewalls and other kinds of barriers.

Our work is available under a 3-clause free BSD source license.  There are no fees or patents required to use the system, and we distribute source code for all the key platform elements.  If you port an application of your own to run on GridCloud, you are not under any obligation to use this same license for your system.  In this sense GridCloud is like an operating system: your application runs on it, and you retain full rights on your system even though our platform is open source.  3-clause BSD does not include any form of "copyleft" restrictive language.

GridCloud downloads:

We do not currently have a single software download for the entire GridCloud system; instead, we have a set of subprojects and you'll need to download them one by one as needed.  This will eventually change, but only when the entire system is fairly stable.   The sub-parts of this project are:

Isis2:  Several of the components of GridCloud are based on software library we call Isis2.  Find it at http://isis2.codeplex.com, with extensive documentation.

CloudMake:    A key task in operating a power grid analysis platform is to implement a mechanism to keep the key elements up and running, 24x7, even when nodes come and go, or crash, loads change, demands on the system evolve, new applications are launched or terminated, etc.  In GridCloud we use a system called CloudMake for this purpose.  CloudMake uses a syntax much like the Linux Makefile syntax but in addition to being able to build software components (Make is normally used to build binaries from source files), CloudMake also is able to sense node and program state through XML formatted files, and when those change, CloudMake will run the associated dependent rules to initiate any needed reconfigurations.  To download CloudMake, visit the URL https://cloudmake.codeplex.com/

Data Collector:    The data collector is a component of GridCloud used to capture incoming PMU or PDC data streams.  It can archive them in files for offline analysis, and/or can relay them to a state estimator for real-time state reconstruction.    https://dc.codeplex.com/ 

GridCloud Example Application: WSU Phasor State Estimator (PSE): The example application used to test the capabilities and robustness of the GridCloud project. The PSE uses the C37.118 PMU data made available via GridCloud to perform state estimations on up to 6,000 electrical busses at up to five times per second. Additional information is located Here.

Live Distributed Objects for Isis2 and GridCloud: The Live Distributed Objects subsystem is used to create a kind of virtual sharable iPad, complete with a display of the power grid or some portion of it, with real-time updates shown as events occur, and with applications ("apps") you can drag and drop onto the grid to perform needed analysis or other kinds of special opertations.  The URL is https://ldo4gridcloud.codeplex.com/

Real Time Snapshot File System: This project is still under development.  It offers a strongly consistent real time file system that can be shared among GridCloud users and will update in milliseconds, always showing the same data to all views.  It includes a way to create a kind of snapshot fixed in time, on which you can run applications that might otherwise be confused if data changes while they are analyzing it.  The Real Time Snapshot File System is optimized to take full advantage of fast RDMA data transfer hardware.

Last edited Apr 27, 2015 at 2:49 AM by daveanderson, version 10