Balancing The Workload In A Cloud Environment | Part I

In a cloud environment, a policy limit is an important desirable attribute – it is used to check and manage resources when the workload demands need to be dynamically balanced after reaching a level of pre-determined threshold.

The policy causes the system to create instances of resources depending on how many workload demands exceed the threshold level.

Before going into more detail on considerations for establishing and using a political boundary, with the goal of balancing the demands of the workload, dynamically create and release resources automatically, we set the policy limit in this context.

Overview of Policy Limit

Let’s take a look at some key attributes of the policy limit.

01. Response times:

The response period between the time the system detects that the workload demands have reached the threshold level and the time that creates the instances of additional resources should be as close as possible to being instantaneous.

When workload demands return to a point below the threshold level, the system deallocates those resources and puts them into another use.

02. Considerations on influence:

The information that must be a limit policy are influenced by factors

The type of cloud service that the consumer rents.
How much control the consumer has on operating systems, hardware and software.
The type of market segment in which the consumer is (e.g., retail, energy and utilities, e-governance services, financial markets, medical and petrochemical).

03. The service provider and policy:

The service provider may be internal cloud in a data center controlled by an organization or be hosted externally by a member of the telecom market segment.

The provider should ensure integration with administrative support functions, such as requests, supply, measurement, classification and collection, billing and other support functions activities and consumer transactions.

04. As the policy limit may be applied:

An example is the case of a consumer segment’s retail market in a cloud service that had a large-scale application in a data center that performed the validation of credit card in the cloud while the workload demands were below the level limit.

When it was time of Diwali shopping, the system detected higher workload demands, exceeding the threshold level. In response, the system quickly created instances of additional demands to balance the workload dynamically.

As the retailer went out of the critical time of purchase, the demands of the workload fell below the threshold level, so that instances of resources in the cloud that were created were released.

As the organization has some control over the hardware, it is able to negotiate with the service provider cloud based on the terms set out in policy. It’s always good to negotiate the parameters of the policy before the critical time of purchase.

The rest of this article provides some background on the types of cloud services and shows how a policy to limit a type of cloud can be different from the policy to another type of cloud.

In addition, it discusses the policy limit on resource management for testing, production and capacity planning applications and notes some of the most important issues to be considered as impacts of a policy to limit a service level agreement (SLA) .

Types of cloud services

First, consider which of these three types of cloud computing services meet your needs:

• Software as a Service (SaaS)
• Platform as a Service (PaaS)
• Infrastructure as a Service (IaaS)

We will also discuss how the size of your operation can influence whether your best option for a type of cloud service is public or private.

01. Software as service:

Suppose, as a consumer market segment of retail, you get a license to a SaaS provider for your company in order to run an application to use the Web as a service on demand.

You choose a subscription or payment method in accordance with use, because you have no hardware or software to buy, install and maintain, no need to update the application.

The only control you have is to use the provider application from a desktop or a remote device, process, business tasks such as billing and computerized billing and human resource management.

Although you do not manage deployed applications, operating systems, storage and networking, you need to know a policy limit on the provider’s resource management if there is a peak, planned or unexpected, demand of workload:

You should know how the provider sets threshold levels to ensure continuous operational availability of SaaS.
Must know what are the terms of the SLA and the backup policy of the provider.
If the service fails because the provider could not handle a surge in demand dynamically, you should know if you can get credits, rebates, free months or terminate the SaaS as determined in an SLA.

02. Platform as a Service:

With PaaS, you can develop retail applications from creation to implementation to application testing (or production as a service).

Unlike SaaS, you can control all the applications located on a lifetime of business integral to the platform. For example, spreadsheets, word processing, backups, recovery, processing payroll and billing.

The provider controls the operating system, hardware or network infrastructure in which applications are running. The provider can develop, implement, execute and manage updates and patches on all the features of, say, a retail management
application.

Of course you want a policy limit of PaaS provider:

You should know how the provider sets threshold levels to ensure that PaaS is still available.
If the service fails because the provider could not handle a surge in demand dynamically, you must obtain credits, rebates, free months or terminate the service.

03. Infrastructure as a service:

With IaaS, you can control the operating systems, network equipment and applications deployed on the virtual machine level:

• You can scale the number of virtual servers or blocks of storage area for up or down.
• You can use the infrastructure pay for these resources in the traditional computing environment in the cloud.

You will need to know a limit policy for the IaaS infrastructure provider:

You should know how the provider sets threshold levels to ensure that IaaS will be maintained when there is a peak in workload demands.
Must be able to negotiate the terms of the policy limit and the SLA with your IaaS provider company.
If the service fails because the infrastructure of computing resources could not handle a surge in demand dynamically, resulting in slow response times, you must obtain credits, rebates, free months or terminate the service as determined in an
SLA.

04. Scale of operations: the public versus the private

As an example, my company generates revenues greater than 1 Crores. We think private clouds may be more economical than public clouds.

A cloud has many private domestic business of the same features that a public cloud, but with much higher rates of e governance, security, availability and resilience of small businesses with revenues of, say, less than 1 Crore.

With a public cloud, data can be stored in unknown locations and cannot be easily retrievable. This is in contrast to a private cloud that allows a consumer to recover data from known locations in a particular jurisdiction.

Unknown locations are not suitable for storing test data for compliance, privacy and confidential. They can be in geographical areas in which privacy regulations and compliance of a country are different from this kind of law in another country. Laws vary from one country to another with respect to export controls data.

When creating a policy limit, my company requires the highest levels of dynamic balance of workload demands in a cloud environment. The system must be able to quickly create additional instances of resources when the workload demands exceed
the threshold level.

Due to the large size of operating my business-oriented workloads per transaction are higher than they would be for small businesses. The range and number of transaction types are higher for small businesses.

As the types of transactions are identified by code number or character of two or three bits, a large company or small business needs to associate a category of business transaction to each type. A category of business transaction suitable for a large company (such as financial leasing) may not be suitable for a small business.