Six steps in migrating to private cloud

Building up a private cloud includes standardization, cable consolidation, storage optimization, automation, and orchestration, according to Logicalis.

With the growing trend of IT departments moving to the cloud, some will favor private clouds over the traditional data center approach, especially for more sensitive applications or in order to comply with industry regulations.

The key steps in building up a private cloud include standardization, cable consolidation, storage optimization, automation, and orchestration, according to Rob Gee, Solutions Architect at Logicalis.

1. Standardize and simplify – reduce cost and increase agility

Read this

Security implications of public vs. private clouds

Nothing it seems is more confusing than the differences between public and private clouds, and beyond that, the safety and security implications of using either type of service. So in this article, we run through some of the basics.

Read More

Virtualize and unify as much of the architecture as possible. This will significantly ease management, often reducing between 20 and 60 points of management down to just 5. Take all the network components and create a single management point, and do the same for storage and compute resources.

2. Cable consolidation – reduce cost and increase utilisation

In this second step, reduce the number of network connections. A typical server has nine cables coming out of it. With the number of servers some businesses have, consolidation will be complex. In the ideal unified network, each server is connected via single cable. Modules/ancillary components and switch ports should also be reduced.

3. Storage efficiency – reduce cost and increase efficiency

In most cases, only 5 percent of a company's stored data is active at any one point. Performing de-duplication can allow the business to reclaim 45 – 95 percent of its capacity. Other steps can also include thin provisioning and storage tiering.

Auto storage tiering will enable an organisation to structure its stored data in an optimal way. Users can also make snapshot copies, thin replications or virtual clones to reclaim significant storage capacity.

4. Automation – reduce cost and decrease time to market

Virtualizing the compute layer effectively transforms the CPU into "stateless" processing power and memory. The processor can then be dynamically assigned to execute the task at hand. Stateless computing effectively does to hardware what hypervisors do to operating systems. This will result in far more rapid provisioning.

5. Uniform scale – reduce cost and provide linear scalability

Determine the capacity needed post the above steps and enjoy the flexibility. Private cloud allows businesses to scale in a uniform manner. As the business grows, they can add RAM or any additional capacity as needed. This removes the need for multiple networks and reduces cable count and switch port count. Of course there will be additional cost incurred to increase capacity, but this is overshadowed by the effectiveness gained.

6. Orchestration and management – spend more time innovating

Orchestration allows quick provision of storage, network, compute and applications from a single pane of glass and enable services such as back up and replication. This will also enable IT to provide charge back and show back; giving a true holistic view of the data centre and applications, as well as reporting on infrastructure utilization.

Show Comments