Most enterprises rely on their data centers to offer services to internal and external users. Important concerns for any data center today include the following:

  • Faster provisioning of new services to meet sudden business demands
  • Scaling existing services without affecting users
  • Optimizing resources for reduced power and energy costs
  • Achieving service-level agreements (SLAs)
  • Reducing management complexities
  • Reducing energy costs

Provisioning and Scaling Services

Data centers must be more agile to meet growing business needs. The conventional infrastructure is very rigid and takes a lot of time for both provisioning and expansion. Each and every service delivered to users involves intense network, server, storage, and application configurations and usually takes months to set up.

Today, however, businesses don’t have the luxury to wait for such long periods. Thanks to converged infrastructure—which integrates critical components of IT such as network devices, servers, and storage devices and provides a single management framework—the result is faster service provisioning and expansion.

Unlike a conventional infrastructure in which all devices are managed as separate entities, a converged infrastructure allows network, servers, and storage to be managed together. All the management configurations are maintained centrally as service profiles. So whenever new services have to be delivered or existing services have to be scaled, existing service profiles are copied and modified accordingly; therefore, the time to deliver new services or scale existing services is drastically reduced. According to IDC, data centers can roll out new services nearly four times faster with converged infrastructure.

Optimizing Resources

Converged infrastructure adopts blade servers and fabrics, which are very small in size and rack mountable yet are equivalent to physical servers and network devices, respectively. Blade servers and fabrics consume very little floor space and power compared with physical devices, and that feature reduces power and energy costs.

Converged infrastructure adopts technologies such as server virtualization, storage virtualization, SDN, and much more. Although these technologies increase the agility of a data center, they also tend to underutilize the hardware resources and result in IT sprawl. By optimizing the resources, data centers can dramatically reduce their hardware, software, and maintenance costs.

Achieving SLAs

SLAs are very important and determine the success of the services delivered. The discrete items such as network, server, and storage are converged into a single physical entity that is easy to manage and far less likely to fail. And when failures do occur, they are much easier to fix in a converged infrastructure, as the admin just has to remove and troubleshoot the blade server that had the failure and insert a new one.

Since converged infrastructure enables rapid provisioning of new services, new networks can be created and the services can be provisioned quickly—even during downtime—by launching the same configuration. Enterprises have increased their service availability from 99.89% to 99.996% using converged infrastructure, according to the IDC report “Measuring the Business Value of Converged Infrastructure in the Data Center.”

Reducing Management Complexities

Converged infrastructure should not be managed in the same way as conventional infrastructures. In a conventional infrastructure, each and every device is managed as a discrete entity. When managing a converged infrastructure, by contrast, the management software should be intelligent enough to understand the same mapping available inside the converged infrastructure and manage the resources accordingly. Only then can it be managed better and optimized for performance.

Visibility into the network, server, and storage layers is very important when managing a converged infrastructure. Apart from monitoring the hardware resources, admins should also be able to manage the network configurations, monitor network traffic, monitor physical and virtual servers, and monitor other hardware such as fans and power supplies.

A clear inventory with the logical mapping between the devices is also important. Physically, all the network, server, and storage devices will be available in a few boxes, and the configuration (service profile) defines the relationship between them.

Fault management is another critical aspect to consider when managing a converged infrastructure. When a fault occurs, admins and technicians must be notified immediately via email or SMS, so they can troubleshoot the fault before it causes a serious problem. Faults with repeated troubleshooting steps should be automated to increase the IT team’s productivity.

Reducing Energy Costs

Constantly rising energy costs are a serious problem for data centers. Up to 60 percent of a typical data center utility bill is attributable to cooling systems. Considering the huge size of today’s data centers, those bills can run into the millions of dollars over years. Converged infrastructure reduces energy consumption, yielding substantial savings on energy costs.

Converged infrastructure devices, however, still need to be monitored for power consumption and temperature. Increased power consumption raises the temperature of the devices and, in turn, requires more cooling from computer-room air conditioning. It is a common practice for devices to be supplied with the amount of power that is noted on their nameplates, but in reality, they consume less than that, so by monitoring the actual power consumption, the amount of excess power being supplied can be identified and blocked.

Converged infrastructure offers numerous benefits such as faster service provisioning, reduced opex and energy costs, and increased service availability. Effective management is necessary to ensure peak performance, however. Data center managers must be thorough when managing converged infrastructure and must address 10 critical areas:

  • Manage components from multiple vendors such as Cisco (UCS), VCE (Vblock), NetApp (FlexPod) and IBM (PureFlex)
  • Monitor all components of a converged infrastructure
  • Manage network configurations
  • Analyze network traffic
  • Monitor physical and virtual servers
  • Map relationship via CMDB
  • Alert for faults via SMS and email
  • Automate repeated IT tasks and L1/L2 troubleshooting
  • Visualize converged infrastructure in 3D models
  • Generate in-depth reports

This blog post was originally published in: