Transforming IT
cancel
Showing results for 
Search instead for 
Did you mean: 

Moving towards Modularity: a crucial milestone on the road to the Converged Data Center

TS_Guest

By James Cohen, Lead Portfolio Architect, HP Data Center Consulting - IT Consolidation & Virtualization

 

James_Cohen_badge.pngMuch has been written and said about how IT organizations can save money through consolidation and virtualization – and if you believe everything that has been said, virtualization is the solution to all IT’s current cost- and agility-related issues. This is one of a number of myths relating to virtualization, similar to the myth that mission critical applications cannot be virtualized (when they almost certainly can) or that the hardware selected for virtualization does not matter, because it is a commodity product – when the decision you make will have implications for your journey to the Converged Cloud.

 

Converged Cloud.pngIf you want to move to a Converged Cloud Solution, whereby the business users of an organization can access solutions from either traditional IT or from the cloud, it is important that the traditional IT infrastructure is optimally designed to deliver services.

 

To deliver traditional IT solutions you need to consider well beyond the conventional servers, storage and network, security and service management.  A successful Converged Data Center requires the right sourcing (your own data center, outsourced, private cloud or public cloud), right sizing of the data center (building, power, cooling, location and security) and the correct delivery organization (people, process and governance.  You need to converge these four domains to deliver a successful Converged Data Center.

 

When it comes to security, customers are used to securing operating systems and access to the infrastructure, but the introduction of the hypervisor also introduces another layer of complexity and vulnerability, which needs to be addressed.

 

To enable a customer to transition in a phased, managed and cost-effective manner to the Converged Cloud, HP TS Consulting has developed a new approach, using a simple model that will deliver all the elements of the Converged Data Center (see the chart below). This then maps onto the services that HP Technical Services can deliver to take the customer on the journey.

 

Convergence_model.png

 

 

Chris Coggrave’s blog Converging the Data Center for a Hybrid Future provides an overview of this approach; I will focus specifically on the Abstraction of the infrastructucture from the services that it delivers.

 

How does Modularity fit into the Converged Data Center?Components of convergence.png

 

There are basically two options within the traditional data center – workloads can be either hosted on physical infrastructure or virtual infrastructure.  Most organizations have a combination of both, usually for legacy reasons.  Many organizations have been reluctant to use virtualization on mission critical applications because in the past applications vendors have been unwilling to support their applications in hypervisor – but this is now historic.  Some organizations have had governance issues to content with, such as business units being reluctant to put workloads on the same physical infrastructure, or concerns about data leakage on physical servers.  Most of these issues are in the past, and organizations should now be looking to drive up their virtualization ratios, particularly if they are looking to improve availability or prepare for cloud services.

 

Virtualization has been the recommended approach to reducing costs and driving up asset utilization for the past few years.  Most application vendors now support their applications on hypervisors; there are well documented benefits related to cost, agility and availability, and I do not intend to look further at business cases in this blog. 

 

However, virtualized servers tend to be large, expensive servers, with virtualization being used to segment the large CPUs into smaller logical units. With the emergence of Extreme Low Energy servers, such as HP Moonshot, there is an argument that small, low cost servers (without the need for the cost and complexity a hypervisor introduces) may, for some workloads, be a more effective solution than a large server (with a hypervisor).

 

So how might one determine what the correct outcome would be for a customer?

 

HP’s approach  is well defined and is focused not just on the server infrastructure, but the entire IT infrastructure, the supporting data center and the needs of the business.

Converged cloud process.png

 

HP’s  Converged Infrastructure Transformation Workshop and Converged Infrastructure Capability Modeling tools (pdf download) help stakeholders quickly agree on the required outcome – from both business and IT viewpoints – and establish close Business-to-IT alignment.

 

The outcome from the workshops will also help identify the required service delivery model, which will help determine where potentially Public Cloud services might be appropriate, the governance requirements for the organization, and any compliance and regulatory framework which must be complied with.

 

The desired future-state business outcome will drive the recommended infrastructure and the optimal combination of physical and virtual infrastructure.  In determining the business case (including the cost of delivering the program) HP’s approach is not about simply looking at the capital cost of a new infrastructure; it’s also about the cost of operation and even the costs of high-availability services and support over the life of the infrastructure.  When it comes to discovery, design and implementation, HP’s tooling is hypervisor-agnostic, and where applicable it automates many of the processes to keep the cost of the project as low as possible.

 

For services that are not going to be outsourced or placed into the public cloud, HP will undertake a Discovery Service that will discover the physical assets within the scope of the discovery, the performance of the assets, and the applications. The level of automated discovery will depend on the tooling selected, the existing infrastructure and any security constraints that may be applied.

With the discovery completed, HP will work with the customer to determine the optimal future state infrastructure to meet the business and IT needs. And in the design phase, we will design the future infrastructure, the future service delivery and service management models, the future facility requirements, and the migration plan for each application and its supporting data.

 

With the planning of all elements of the modularization of the infrastructure completed, and accepted, we can move forward into the implementation phase, where the design is implemented, and the migration to the future state completed.  Included in this phase is not only the implementation of the physical infrastructure, but also the migration of the applications to the new infrastructures (whether internally delivered or through some sort of outsourced/cloud solution), and the delivery of the new organizational model, with the recommend tooling and processes.

 

Once in production, HP can support the new solution through Data Center Care services for any infrastructure, and health check services to ensure that the solution continues to deliver the business benefits that were identified at the start of the engagement.

 

To learn more about HP’s comprehensive products and expertise for the Converged Data Center, see this short video by Chris Coggrave.

 

To learn more about me and how I can help you build a cost-effective data center solution that meets your business needs, visit my HP Technology Expert Profile.

 

 

0 Kudos
About the Author

TS_Guest

Labels
Events
June 18 - 20
Las Vegas, NV
HPE Discover 2019 Las Vegas
Learn about all things Discover 2019 in  Las Vegas, Nevada, June 18-20, 2019
Read more
Read for dates
HPE at 2019 Technology Events
Learn about the technology events where Hewlett Packard Enterprise will have a presence in 2019.
Read more
View all