Documentation
The Grid Computing Service offers multiple methods for accessing it and processing computations on the machines managed by it. In order to understand the best way to manage your job submissions, some information about how the Grid Service is set up is necessary.
First of all, in order to access the grid, you will need an account with the service. Once given an account by an administrator, depending on what access your account has been granted, you can create new applications, submit jobs to be processed and retrieve data from completed jobs.
There's a few ways that this can be accomplished. One simple way to access the grid is through the Grid Console. More commonly, users will be accessing the Grid Service through existing command-line interfaces or through custom applications written based upon the United Devices software development kit (SDK). Your method for submitting and retrieving jobs will depend on what kind of data processing you will be doing and which application you will be using.
How the Grid Works:
Jobs are submitted by you to the Grid Computing Service. Depending on what machines you have access to and what kind of application you are using for your computations, the Grid Service will distribute your job to client machines that are attached to the Grid. These client machines will compute a portion of your data and send the results back to the Grid Service. Once all of the portions of data have been computed and returned, the Grid Service will reassemble the computations and store the results in a database for when you are ready to retrieve them.
Applications:
An Application is a suite of programs that are common in their intended task. The application defines how the Grid Service will interact with a program, how the program will access a data set and what kind of results are expected to be returned.
Programs:
A Program is an executable piece of code that does the actual computations. The Grid Service distributes this code to the client machines where the actual processing of your jobs are completed.
Data Sets:
A Data Set is a piece of data that the program runs it's computations on. These can be in many forms including databases, text files, binary files and more.
Jobs:
A Job is a command sent by you to the Grid Service. It includes which application group you intend on using, the program within that group that you want to use and data sets for which you need computations performed upon.
