Resource Management Models in Cloud Computing

Pre-requisite:- Cloud Computing

The term resource management refers to the operations used to control how capabilities provided by Cloud resources and services are made available to other entities, whether users, applications, or services.

Types of Resources

Physical Resource: Computer, disk, database, network, etc.

Logical Resource: Execution, monitoring, and application to communicate

Resource Management in Cloud Computing Environment

On the Cloud Vendor’s View

• Provision resources on an on-demand basis.
• Energy conservation and proper utilization is maintained in Cloud Data Centers 

On the Cloud Service Provider’s View

• To make available the best performance resources at the cheapest cost.
• QoS (Quality of Service) to their cloud users

On the Cloud User’s View

• Renting resources at a low price without compromising performance
• Cloud provider guarantees to provide a minimum level of service to the user

Resource Management Models

Compute Model

Resource in the cloud is shared by all users at the same time. It allows the user to reserve the VM's memory to ensure that the memory size requested by the VM is always available to operate locally on clouds with a good enough level of QoS (Quality of Service) being delivered to the end user. 

Grid Strictly manages the workload of computing mode. Local resource manager such as Portable Batch System, Condor, and Sun Grid Engine manages the compute resource for the Grid site. Identify the user to run the job

Data Model

It is related to plotting, separating, querying, transferring, caching, and replicating data.

• Data is Stored at an Un-Trusted Host:- Although may not seem the best policy to store data and let others use the data without permission moving data off-premises increases the number of potential security risks.
• Data Replication over Large Areas:- Making sure data is available and durable whenever demanded is of utmost importance for cloud storage providers. Data availability and durability are typically achieved through under-the-covers replication i.e., data is automatically replicated without customer interference or requests.
• Problems with Data Management:- Transactional data management is one of the biggest data management problems. It is hard to ensure Atomicity, Consistency, Isolation, and Durability is maintained during data replication over large distances. It is also risky to store such sensitive data in untrusted storage.

Virtualization

It is the method by which we can create an emulation of software or hardware on our computer. It has further two components:-

• Abstraction:- Provides the necessary virtual versions of raw compute, storage, and network that can be unified as a pool of resources and resource overlay which includes data storage services, and a web hosting environment.
• Encapsulation:- A virtual machine can be represented as a single file. Virtualization configures, deploys, starts, migrates, suspends, resumes, and stops in each application. Provides better security, manageability, and isolation.

Monitoring

The challenge that virtualization brings is that users don't have a lot of control over the monitoring resource. It is a method of reviewing, observing, and managing the operation of a cloud-based IT infrastructure.

• In Cloud: Different levels of services can be offered to end users. The user is only exposed to a limited Application Programming Interface. And lower-level resources are not revealed to the user (PaaS, SaaS level some providers may choose to expose monitoring information at these levels). The user does not have the capability to implement a new application of its own monitoring infrastructure. Limited information returned to users restricts their knowledge about the current status of the resource. Require to maintain business tracking, update, inspect and troubleshoot the servers of the cloud organization, monitor virtual machines, and maintain the functioning of the hardware.
• In Grid: Have a different trust model in which users via their identity delegation can access and browse resources at different Grid sites and Grid resources are not highly abstracted and virtualized as in Clouds.

Programming Model

User-level programming languages are used for accessing and operating the cloud.

• In Cloud: Makes use of Web Services where users have more control over the Cloud Services. The translation of data for the receiving system and real-time data exchange between systems without middleware of all the services and applications remain a big challenge.
• In Grid: Makes use of parallel and distributed computing environment

Challenges:

1. Multiple service providers allow to access data to clients with little authorization or authentication
2. Diversity in resources in turn affects the performance and stability
3. Error handling in a continuously changing business environment  

Security Model

Allows users to control the security of their own data by maintaining passwords, and receiving any news regarding suspicious activity with their data via email.

Risks in the Security Model:-

1. Privileged use access
2. Regulatory compliance
3. Data location
4. Data partition
5. Recovery
6. Investigation support
7. Long-term durability