Server virtualization and cloud computing are common parlance in the IT industry, but the
concept of private cloud architecture is relatively unclear and fraught with confusion. So I would
say that the shortest definition can be, 'Virtualization on top of a high performance computing
(HPC) platform'.
The fabric of virtualization has its limits. For example, there is little flexibility in
scaling a virtual machine. The scaling will be always limited to the maximum resources available on
a particular physical server. For instance on a 24 core server with 32 GB RAM, you cannot have a
virtual machine which has more than (or even equal to) 24 cores and 32 GB of RAM.
However, if there are 50 such servers, then you technically have 1200 Cores and 1.6 TB of RAM. But
if an organization needs a machine with higher specs from its virtual infrastructure, say 50 cores
and 64 GB of RAM, it's just not possible. The organization has to buy a new server with more than
50 cores and 50 GB of RAM. This can be used as a physical box or to create a virtual machine.
With infrastructure as a service (IaaS), an organization first aggregates its data center as a
single resource pool, and then starts provisioning virtual machines on it. This is similar to
creating an HPC platform which integrates all your existing resources in a single pool and then
running a virtualization layer on top of it.
Very few companies provide full-fledged private
cloud architecture.
Requires Membership to View
To gain access to this and all member only content, please provide the following information:
By submitting your registration information to searchDataCenter.in you agree to receive email communications from the TechTarget network of sites, and/or third party content providers that have relationships with TechTarget, based on your topic interests and activity, including updates on new content, event notifications, new site launches and market research surveys. Please verify all information and selections above. You may unsubscribe at any time from one or more of the services you have selected by editing your profile, unsubscribing via email or by contacting us here
- Your use of searchDataCenter.in is governed by our Terms of Use
- We designed our Privacy Policy to provide you with important disclosures about how we collect and use your registration and other information. We encourage you to read the Privacy Policy, and to use it to help make informed decisions.
- If you reside outside of the United States, by submitting this registration information you consent to having your personal data transferred to and processed in the United States.
One such commercial offering is Enomaly, while Eucalyptus is an open source
variant that has gained attention. You have to keep a few things in mind while deploying private
cloud architecture:Server architecture: This technology runs on the Hypervisor-based virtualization. Servers need Intel VT or AMD Pacifica features to run as part of the private cloud architecture. So it's essential; that you rule out old servers.
Selection of cloud middleware: It's best to opt for a private cloud
architecture supported by your existing server virtualization platform. A XEN user can select
between Ubuntu Enterprise Cloud or Enomaly. If you are a VMware fan, then sticking with VMware
based private cloud architecture is a better option. Microsoft Hyper-V users will have to wait for
the time being. Microsoft has promised to launch its Dynamic Datacenter Toolkit, which will let
users build private
cloud architectures based on Hyper-V.
It is important to assess whether you require a hybrid cloud, and whether your cloud platform
supports public
cloud architecture. Eucalyptus which runs inside the heart of Ubuntu Enterprise Cloud is Amazon
EC2 compliant, and can connect with EC2 to form a hybrid cloud. Similarly, we can speculate that
Microsoft's Dynamic Datacenter Toolkit will integrate well with Microsoft Azure.
About the author: Anindya Roy is a tech evangelist at Netweb Technologies, and has worked on
open source private cloud platforms. His specialization areas include HPC, cloud computing, and
virtualization. Roy also has an extensive experience in application evaluation as well as
testing.
(As told to Jasmine Desai.)
This was first published in June 2010