ORLANDO -- The recession of the last decade led to enormous cost-saving projects as businesses tightened belts...
and weathered the economic storm, with projects including virtualization and consolidation, exploration of fledgling cloud technologies, centralizing data centers, rethinking the way software is developed and more.
At Gartner's IT Operations Strategies and Solutions Summit 2015 here this week, vice president and distinguished analyst Jay Pultz explained the time is right to consider another round of IT infrastructure and operations cost cuts. He outlined six major areas where organizations can find potential savings -- though investments may be required to realize long-term savings.
1. Software-defined and integrated infrastructures
Pultz started his cost-saving recommendations where 42% of infrastructure and operations (I&O) expenses are estimated to reside. Two technologies come immediately to the forefront: software-defined data centers and integrated systems, more commonly referred to as converged infrastructures (CI).
The move to adopt software-defined data centers (SDDC) is recognition of the flexibility and efficiency to be realized when software is used to provision and manage compute resources that are pooled together from across the data center. It eliminates traditional IT silos and forces businesses to re-think server, network and storage administration as a more holistic pursuit.
Although SDDC and CI both require capital investments to implement, Pultz is confident of the value after running numerous analyses for clients.
"It turns out that the combination of software-defined data centers and integrated systems is significantly cheaper on a [total cost of ownership] TCO basis," he said. Where SDDC brings flexibility, integrated systems can guarantee systems integration and shrink the physical footprint with less space, fewer boxes, lower maintenance expenses and easier service.
2. Server modernization
Pultz makes an economic argument for replacing aging servers with current models that are better able to support high levels of workload consolidation (lots of VMs). Organizations that plan on high volumes of servers might also consider commoditized servers such as Open Compute Project designs. Commoditized servers can be custom-built and are basically intended as throw-away devices -- there are no maintenance agreements or expenses -- which lowers the server TCO further.
3. Network negotiations
Networks can consume 26% of an organization's I&O budget, with the majority of those costs telecom services that provide the connectivity between your local- and wide-area networks (LAN and WAN). Pultz offered a range of advice for addressing telecom expenses.
Prices for carrier services such as 4G are falling; organizations typically buy far more bandwidth than they actually need and it's common to see contracts with multiple providers. So Pultz recommends a comprehensive re-evaluation of providers, costs, services and deals to pare back unnecessary expenditures. It's worth making the effort to re-negotiate telecom carrier contracts to more favorable terms -- especially if your current agreement is up for renewal.
Consider wireless rather than wired LAN deployments where possible or practical. Wireless LANs are easier to manage and don't require the physical effort to move or patch physical cabling each time a change occurs in the environment. If you're engaging a colocation provider with its own network connectivity, it might be worth utilizing their network for at least some of your bandwidth needs rather than investing in additional telecom services or equipment locally.
4. Assess client device management
Client computing can consume 19% of an I&O budget, so Pultz suggests organizations with lots of local PCs take a fresh look at the way endpoints are managed. For example, traditional PCs can be extremely costly to manage and maintain due to tasks such as applying different patches to different software versions on each PC. Making the shift to thinner standardized systems such as tablet/laptop hybrids can save purchase costs -- and managing those endpoints with effective tools can reduce diagnostic/troubleshooting time and configuration errors while bolstering security.
Organizations that seek the pinnacle of endpoint uniformity and manageability should consider desktop virtualization on thin client systems. However, VDI imposes significant data center resource demands and is best for businesses with large numbers of users with identical needs, such as call centers.
5. Self-service desks
Although service (help) desk operations normally demand only 13% of a company's I&O budget, Pultz notes that 85% of that cost is consumed by the staff; yet 60% of the calls service desks field are routine. This makes self-service resources extremely attractive as a cost-saving strategy. For example, users who can perform basic tasks like resetting their own passwords or recovering deleted emails can reduce service desk calls and save the business significant IT infrastructure costs on the staffing side.
6. Consider the cloud
Finally, no discussion on IT infrastructure cost reduction would be complete without including the role of cloud computing in the modern enterprise. Pultz says that cloud resources like Amazon Web Services EC2 server instances can reduce I&O costs by 47% over three years, while S3 storage instances can save 79% of I&O costs over the same period.
Pultz underscores such raw savings will be tempered by the cost of people who would still be needed to make cloud deployment decisions and perform monitoring, software development and so on. In addition, cloud resilience is improving and the range of applications that can reliably live in a public cloud is growing. But mission-critical applications are still best kept in-house where the IT staff can exert direct control over their security and availability.
Trends to follow through the data center
Craft a strong IT infrastructure budget