Advanced Distributed System - Resource Management

What are resources?

All stuffs needed for maintaining the run of a system are called resources, such as hardware(CPU, storage, network links), software(web services), and datasets. Applications need to share the resources, and we need to consider how to optimize their use with respect to resources available and resources requested by different applications.

One example demonstrating a trade-off between time, price and storage resource is that, we could make a software parallelisable and run it with different expensive processors, but it uses lots of storage because large amount of data are processing at the same time. Or, run the software sequentially on a cheaper processor with less storage used because smaller amount of data is active at the same time.

Grid Computing

One of the definition of grid computing is: Resource sharing and coordinated problem solving in dynamic virtual organizations. The grid computing coordinates the resources not subjected to centralized control, it use standard, open, general-purpose protocols and interfaces to deliver service. Some functions of grid computing are:

• Resource discovery and information collection and publishing
• Data Management on and between resources
• Process Management on and between resources
• Common Security Mechanisms
• Process and Session Recording/Accounting

Cloud Computing

The idea of cloud computing is, by providing resources as a service to users who don’t have their own resources. This is called on-demand resource provisioning. It relies on virtualization of resources that create virtual machines as opposed to physical machines.

There are some cloud computing service models:

1. Software as a Service (SaaS):
   Provides software as services. E.g. Microsoft 365, Dropbox
2. Platform as a Service (PaaS):
   Provides environment of programming functionalities. E.g. Google Compute Engine
3. Infrastructure as a Service (IaaS):
   Provides pay-as-you-go access to storage, networking resourses and others. E.g. Hadoop + Mapreduce

And in cloud computing, there are two roles - the cloud provider and cloud consumer, where a cloud provider’s objective is to provide services to user and earn profits, and a cloud consumer wants to buy resources and apply them to solve tasks. On the perspective of a cloud provider, there are some aspects need to consider in a question - how to allocate VM(Virtual Machine) to PM(Physical Machine)?

1. How to map services level agreement with users
2. How to maximise resource allocation
3. How to minimise energy cost

• ← Previous Post
• Next Post →