Cloud computing has really reached a period of incredible growth that compelled many IT organizations to increasingly take on a service provider role within their enterprises, with many transforming from cost centers to business units. Cloud computing provides such flexibility that made many IT organizations take a closer look at adopting cloud technology.
The silent shift brings with it heightened accountability to internal customers and a consequent move away from the mere management of hardware toward the delivery of accurately sized workloads to meet business units' needs. This "provide only what's needed" model calls for flexibility at the resource end.
Designed to deliver software, infrastructure and platform resources as on-demand services from a shared pool with a high degree of efficiency, cloud computing is inherently flexible. This enables IT organizations — and even users themselves — to provide/requisition the exact computing power and capacity needed to dynamically support products and services.
The potential rewards are great. Besides enabling IT organizations to fulfill their new charter as a provider of high-quality services, cloud computing can help enterprises reduce IT infrastructure maintenance costs (in some cases, to less than 50 percent of IT spending) while enhancing organizational agility.
Moreover, having a cloud infrastructure also allows enterprises to leverage third-party cloud services such as hosted software or offsite disaster recovery with nil or minimal capital investment.
Understanding IT infrastructure options is the critical first step of any journey into the cloud. Cloud technology is necessarily flexible so one size definitely does not fit all.
Deployment Model, Services
The first thing CIOs and their teams need to decide is which cloud deployment model to use. There are three to choose from — private, public, and hybrid — and each has its own advantages, drawbacks, and potential usage scenarios.
A private cloud is designed for the exclusive use of a single organization and can be operated and hosted either by the enterprise itself or by an external provider. A public cloud, on the other hand, is used by multiple organizations (called tenants) that lease cloud space from a cloud provider. The hybrid cloud model refers to when an enterprise turns to a public cloud for additional processing capacity when its own private cloud is stretched to the limit.
The next thing that needs to be decided is the services that the enterprise wants its cloud to deliver. Three types of services are commonly delivered by clouds: software (e-mail, data backup, etc.), infrastructure (virtual servers, virtual clients, remote storage, etc.), and platform (application development, deployment and hosting).
After deciding which cloud deployment model to use, choosing the combination of the services to be delivered, considering the impact of cloud computing on all areas of the business, and establishing clear goals, the enterprise can then come up with a desktop-to-data center cloud infrastructure that meets its particular needs.
Infrastructure Design
When designing cloud infrastructure, CIOs need to bear in mind that cloud success and efficiency improves when the correct building blocks, integrated management tools, and service delivery mechanisms are used.
Cloud-enabled data centers are distinctly different from traditional ones, with high performance, maximum hardware density, and high scalability among the critical considerations.
Accordingly, standardized cloud-optimized servers with a dense, energy-efficient design are needed at the core, not general-purpose servers used in traditional data centers.
And because high availability in cloud environments is typically achieved in the software layer, many system features and components that are often found in general-purpose servers are not needed in the servers that form the heart of cloud-enabled data centers.
As for setting up cloud infrastructure, CIOs can take one or both of two implementation approaches: revolutionary or evolutionary. The choice depends on the number of traditional enterprise applications that need to be accommodated in the new cloud regime, how supportive existing infrastructural assets are of cloud computing, and on the benefits the enterprise wants to achieve.
The revolutionary approach is ideal for those building greenfield data centers or who have no need to accommodate traditional enterprise applications.
Its evolutionary cousin, which involves incremental migration toward cloud computing, is more suited for those with existing traditional enterprise applications and that want to build on infrastructure they already have.
Today, the evolutionary approach is the predominant approach but as more and more cloud-native applications are developed, the balance is expected to tip in favor of the revolutionary approach.
The silent shift brings with it heightened accountability to internal customers and a consequent move away from the mere management of hardware toward the delivery of accurately sized workloads to meet business units' needs. This "provide only what's needed" model calls for flexibility at the resource end.
Designed to deliver software, infrastructure and platform resources as on-demand services from a shared pool with a high degree of efficiency, cloud computing is inherently flexible. This enables IT organizations — and even users themselves — to provide/requisition the exact computing power and capacity needed to dynamically support products and services.
The potential rewards are great. Besides enabling IT organizations to fulfill their new charter as a provider of high-quality services, cloud computing can help enterprises reduce IT infrastructure maintenance costs (in some cases, to less than 50 percent of IT spending) while enhancing organizational agility.
Moreover, having a cloud infrastructure also allows enterprises to leverage third-party cloud services such as hosted software or offsite disaster recovery with nil or minimal capital investment.
Understanding IT infrastructure options is the critical first step of any journey into the cloud. Cloud technology is necessarily flexible so one size definitely does not fit all.
Deployment Model, Services
The first thing CIOs and their teams need to decide is which cloud deployment model to use. There are three to choose from — private, public, and hybrid — and each has its own advantages, drawbacks, and potential usage scenarios.
A private cloud is designed for the exclusive use of a single organization and can be operated and hosted either by the enterprise itself or by an external provider. A public cloud, on the other hand, is used by multiple organizations (called tenants) that lease cloud space from a cloud provider. The hybrid cloud model refers to when an enterprise turns to a public cloud for additional processing capacity when its own private cloud is stretched to the limit.
The next thing that needs to be decided is the services that the enterprise wants its cloud to deliver. Three types of services are commonly delivered by clouds: software (e-mail, data backup, etc.), infrastructure (virtual servers, virtual clients, remote storage, etc.), and platform (application development, deployment and hosting).
After deciding which cloud deployment model to use, choosing the combination of the services to be delivered, considering the impact of cloud computing on all areas of the business, and establishing clear goals, the enterprise can then come up with a desktop-to-data center cloud infrastructure that meets its particular needs.
Infrastructure Design
When designing cloud infrastructure, CIOs need to bear in mind that cloud success and efficiency improves when the correct building blocks, integrated management tools, and service delivery mechanisms are used.
Cloud-enabled data centers are distinctly different from traditional ones, with high performance, maximum hardware density, and high scalability among the critical considerations.
Accordingly, standardized cloud-optimized servers with a dense, energy-efficient design are needed at the core, not general-purpose servers used in traditional data centers.
And because high availability in cloud environments is typically achieved in the software layer, many system features and components that are often found in general-purpose servers are not needed in the servers that form the heart of cloud-enabled data centers.
As for setting up cloud infrastructure, CIOs can take one or both of two implementation approaches: revolutionary or evolutionary. The choice depends on the number of traditional enterprise applications that need to be accommodated in the new cloud regime, how supportive existing infrastructural assets are of cloud computing, and on the benefits the enterprise wants to achieve.
The revolutionary approach is ideal for those building greenfield data centers or who have no need to accommodate traditional enterprise applications.
Its evolutionary cousin, which involves incremental migration toward cloud computing, is more suited for those with existing traditional enterprise applications and that want to build on infrastructure they already have.
Today, the evolutionary approach is the predominant approach but as more and more cloud-native applications are developed, the balance is expected to tip in favor of the revolutionary approach.
Posts
0 comments
Post a Comment