A private cloud is many different things to many different people. Some stop at virtualization, other take it at the automation of the provisioning and elasticity, and some others take it to the final level and talk about chargebacks.
In our view, a private cloud is a computing environment which provides hosted services to a limited subscriber base, generally within a single enterprise behind a firewall.
Private clouds are a conservative’s answer to a regular public cloud solution. Instead of losing sleep worrying about control, safety and costs of hosting data at a third party vendor’s data center, enterprises build their very own in-company, and mostly in-premises private cloud solution.
A lot of people confuse private clouds with virtualization. Although virtualization of your infrastructure is an essential part of building a private cloud, it takes more for having a private cloud. A cloud takes a layered approach to implementation, wherein virtualization sits at the bottom of the pyramid, and then there is an orchestration layer, and an above this is a portal, allowing the management of resources.
The chargeback metering mechanisms are an added plus for the higher management, for they provide a relatively accurate picture of the “cost of IT” in the organization.
Key elements of a private cloud offering include –
Self Service portal
A web application providing access to the infrastructure/services, which facilitates monitoring and consumption of IT services.
Policy based controls
A full featured and standardized rules engine/framework governing the IT resource consumption.
Standardized Hardware
The hardware (servers and other resources) on the network should be under centralized control and standardized. That would reduce the overall maintenance variability.
Automated Deployment and Maintenance
The deployment of software/patching etc should be automated to reduce manual intervention and increase efficiency. The total manual administration effort should come down to a minimum, reducing cost.
Targeted Independence
There should be very lose coupling across services, effectively allowing services across different environments based on different parameters, transparently.
Elasticity
The ability to scale up or down the resources based on demand. This actually is a very commonly sought after feature.
Automated Infrastructure Management and Maintenance
The infrastructure management and maintenance should be automated, reducing the overall manual dependence and increasing efficiency of the overall system.
These items listed above are not really axioms which HAVE to be followed in order to justify an enterprise installation to be a private cloud; however the more of the above list a company follows, a more mature and stable environment is promised.
If you enjoyed this article, please consider sharing it!
Free Whitepaper
The CIO's New Guide to Design of Global IT Infrastructure
Learn how to be prepared to adapt your environment in a way that supports distributed employees, anytime anywhere collaboration and the need for business continuity during a disaster. Is it possible to eliminate the impact of distance? This paper explores the 5 key principles successful CIOs are using to redesign IT infrastructure of any size. Learn how to be prepared to adapt your environment in a way that supports distributed employees, anytime anywhere collaboration and the need for business continuity during a disaster.Free Whitepaper
The Data Center Build-or-Buy Decision: 6 Key Factors You Should Consider
The need for physical data center capacity continues to grow. For some businesses, the driver is expansion into new markets or geographies. For others, it's the need to deal with growing amounts of data generated by applications with high-capacity demands, evolving end-user abilities, or regulatory bodies that demand ever-increasing quantities of meticulous documentation. If your data center is running out of space or power you have two options. You can build and operate a new facility, or you can lease the capacity you need from one of a growing number of colocation providers. The "build-or-buy" decision between construction and colocation should be weighed carefully, as the choice will affect your company and your bottom line quite literally for decades.
Free Whitepaper
The CIO's New Guide to Design of Global IT Infrastructure
Learn how to be prepared to adapt your environment in a way that supports distributed employees, anytime anywhere collaboration and the need for business continuity during a disaster. Is it possible to eliminate the impact of distance? This paper explores the 5 key principles successful CIOs are using to redesign IT infrastructure of any size. Learn how to be prepared to adapt your environment in a way that supports distributed employees, anytime anywhere collaboration and the need for business continuity during a disaster.Free Whitepaper
The Data Center Build-or-Buy Decision: 6 Key Factors You Should Consider
The need for physical data center capacity continues to grow. For some businesses, the driver is expansion into new markets or geographies. For others, it's the need to deal with growing amounts of data generated by applications with high-capacity demands, evolving end-user abilities, or regulatory bodies that demand ever-increasing quantities of meticulous documentation. If your data center is running out of space or power you have two options. You can build and operate a new facility, or you can lease the capacity you need from one of a growing number of colocation providers. The "build-or-buy" decision between construction and colocation should be weighed carefully, as the choice will affect your company and your bottom line quite literally for decades.
Pingback: Selective routing in Amazon VPC | kuliza