Grid Infrastructure

 

Grid computing is a distributed system architecture that enables the sharing and coordination of computing resources across multiple organizations, locations, and domains. Globus is a widely used middleware for grid computing that provides a set of tools and services for managing and sharing distributed resources. In this blog post, we will explore the two-level infrastructure model used in Globus and how it can be used to build a grid infrastructure.

Two-Level Infrastructure Model

The two-level infrastructure model is a key component of the Globus middleware. This model separates the resource management and job execution functions into two levels:




 

Level 1: Resource Management

The first level of the infrastructure is responsible for resource management. This level includes services that provide resource discovery, authentication, authorization, and data management. Resource management services are typically provided by a resource manager that maintains a directory of available resources and their attributes, such as computing power, memory, and storage.

The Globus Toolkit provides a set of resource management services, including the Globus Resource Allocation Manager (GRAM) and the Globus Resource Specification Language (RSL).

Level 2: Job Execution

The second level of the infrastructure is responsible for job execution. This level includes services that provide job submission, monitoring, and execution. Job execution services are typically provided by a job manager that interacts with the resource manager to select and allocate resources for job execution.

The Globus Toolkit provides a set of job execution services, including the Globus Job Manager (GJMS) and the Globus Job Submission Description Language (JSDL).

Cloud computing Grid computing Distributed computing Computer, cloud computing, text, computer, computing png thumbnail

Building a Grid Infrastructure using Globus

To build a grid infrastructure using Globus, you will need to follow the following steps:

1. Install and Configure the Globus Toolkit

The first step is to install and configure the Globus Toolkit on your computing resources. The Globus Toolkit provides a set of tools and services for managing and sharing distributed resources.

2. Set Up Resource Management Services

The next step is to set up resource management services using the Globus Resource Allocation Manager (GRAM) and the Globus Resource Specification Language (RSL). This involves configuring the resource manager to maintain a directory of available resources and their attributes, such as computing power, memory, and storage.

3. Set Up Job Execution Services

The next step is to set up job execution services using the Globus Job Manager (GJMS) and the Globus Job Submission Description Language (JSDL). This involves configuring the job manager to interact with the resource manager to select and allocate resources for job execution.

4. Submit and Monitor Jobs

Once the infrastructure is set up, you can submit and monitor jobs using the Globus Job Submission Description Language (JSDL). This involves describing the job requirements and submitting the job to the job manager for execution. You can monitor the status of the job using the Globus Job Manager (GJMS) and the Globus Resource Allocation Manager (GRAM).

Benefits of Two-Level Infrastructure Model

The two-level infrastructure model used in Globus offers several benefits, including:

1. Separation of Concerns

The two-level infrastructure model separates the resource management and job execution functions into two levels, enabling a clear separation of concerns and greater flexibility in building and managing grid infrastructure.

2. Scalability

The two-level infrastructure model is designed to be scalable, meaning it can accommodate a growing number of users and resources without compromising performance.

3. Customization

The two-level infrastructure model enables organizations to customize their grid infrastructure to meet their specific needs and requirements.

Conclusion

Grid infrastructure is a powerful distributed system architecture that enables the sharing and coordination of computing resources across multiple organizations, locations, and domains. The two-level infrastructure model used in Globus provides a flexible and scalable framework.

Post a Comment

Previous Post Next Post