Evolution of the Datacenter

Traditional datacenter architectures employ discrete tiers of compute (servers) and storage, which are connected by a dedicated network such as a storage area network (SAN) or a network attached storage (NAS) system.

While SAN and NAS-based architectures have been the mainstay of many enterprise datacenters, they are straining under the demands of a modern virtualized environment. They simply were not designed to handle the explosion of VMs and the resulting I/O needs of today’s enterprise.

While SANs have served admirably, they are quickly becoming a handicap for the next generation of datacenters. SANs have:

  • Poor scalability for virtualized workloads
  • Lack of IT agility. VM provisioning tightly coupled with inflexible back-end storage systems
  • Network performance bottlenecks under the explosion of I/O requirements in VMs
  • Prohibitive capital expenditures on oversized storage arrays
  • Unsustainable operational costs to support SAN fabrics built from multiple vendors

A Radical Change is in Order

Fortunately, leading datacenter architects have designed a better approach. Google, Amazon, Facebook, and Microsoft have embraced converged architectures that collapse disparate compute and storage tiers into unified systems. These completely eliminate the need for network-based storage systems like the SAN.

IT organizations are moving toward data center convergence, forcing storage architects to
re-evaluate SAN protocol selections, storage processes, and best practices.

- Gartner, November 1, 2012

Source: 2013 Planning Guide: Data Center, Infrastructure,
Operations, Private Cloud and Desktop Transformation


The infrastructure convergence pioneered by these trailblazing companies preserves the benefits of shared storage and reduces the cost and complexity of the storage network. It greatly increases overall datacenter performance, scalability, and efficiency.

Turnkey solutions decrease IT cost


Converged solutions tightly integrate compute (servers) and storage resources into a single platform. Because converged systems can be built entirely with off-the-shelf, reliable hardware components, they are cost-effective.

Partnering with a single vendor for a converged design frees you from needing multiple experts for configuration, management, and maintenance tasks. Minimizing that overhead lowers operational costs. Purchasing fewer systems decreases your datacenter footprint, which reduces power usage and cooling expenses.

These simplified architectures reduce costs across the board and accelerate the time-to-value for all virtualized applications. And, turnkey solutions enable datacenter build-outs that are dramatically faster than what’s possible with legacy architectures.

Flexible scalability supports business growth


Converged solutions enable datacenter managers to add scale and performance exactly where it’s needed. IT is able to scale compute resources, storage capacity, and storage performance independently. So, you can build a customized infrastructure that’s optimized for your business needs.

Scaling on-demand means that you don’t have to over-provision the storage array to ensure that you can meet requirements in the future. Convergence lets you deploy the infrastructure that you need today and boost your compute and storage resources incrementally as your business grows.

Convergence eliminates expensive forklift upgrades of aging SAN
or NAS storage infrastructure just to get more performance.
Converged solutions allow linear growth, and they scale easily.

Software-driven agility accelerates time-to-value


A virtual datacenter must be agile, able to provide new services on-demand to keep pace with business needs. Vendors of legacy hardware-centric systems often cannot release new features quickly because their architectures depend on custom ASICs or other specialized hardware.

IT teams that avoid investments in hardware-centric technologies can optimize time-to-value for new datacenter services.

Google, Amazon, and other innovators now rely on software-driven features that run on commodity hardware. That approach breaks the dependency on proprietary hardware. It reduces hardware costs and enables a datacenter to deploy new capabilities rapidly.

A distributed-software architecture that’s well-designed enables a company to scale its operations smoothly and quickly. Software-based solutions have no single point of failure. Your systems grow simply with your business.

Consolidated management streamlines operations


Products that combine compute and storage into a single, integrated system let you manage resources from a single interface. Enterprise IT managers can streamline datacenter operations, ensure reliability, simplify troubleshooting, and monitor the virtual environment through at-a-glance dashboards.

Consolidated management of compute and storage is straightforward, so it’s simple to manage a large-scale virtualized datacenter. Well-designed converged systems preserve the ability of the virtual infrastructure management team to continue to use their existing tools for VM provisioning and management, and not require additional training for new management frameworks. Such an approach allows you to bring new applications online quickly and dramatically reduce operating expenses.

Predictable price and performance drive reliable ROI


Converged solutions deliver server and storage resources that are preconfigured to work together. That setup minimizes the risks of building and scaling an enterprise datacenter, and creates a predictable cost model. Datacenter managers can predict future build-out expenses, so they can estimate return on investment (ROI) with confidence.

Intelligent, distributed software architectures ensure that performance can scale in a linear fashion. Performance does not degrade with the increase of compute demands, storage IOPS, or storage capacity.

As your company grows, you won’t have to estimate individual costs for servers, network switches, SAN gateways, storage arrays, and server-based flash from multiple vendors. Instead, convergence offers a simple IT cost model that allows you to predict future expansion costs.