Fields of Fiber

    Finding Your Way Through the Infrastructure Bottleneck

    Posted by Anthea Strezze on Jan 30, 2017 7:07:00 AM
    Author Image
    Find me on:

    Unsplash_7nrsvjvalna-denys-nevozhai_1200Crop.jpg

    The Digital Transformation Road Trip is now an ebook! Click here to download today.

    This week, we’ll take a look at another impediment to effective digital transformation – the infrastructure bottleneck.

    On the technology side, the biggest source of resistance to change is the physical IT infrastructure itself. IT infrastructure provides the framework that supports all digital transformation. Your legacy infrastructure was most likely architected anywhere from five to twenty years ago, which means that it is rigid and difficult to change. Your network managers are probably already struggling to keep up with the growing demands for bandwidth and agility within your organization and customer base. At the time, the architects who designed your legacy infrastructure probably couldn't imagine the scale of current networks, let alone what we will need in the next few years!

    Realistically, today's predictions of our future needs are also guesses, and they may wind up being just as inadequate. Five or ten years from now, network architects may look back and laugh at the fact that 10 GBPS used to be considered fast. That’s why incorporating software control into the new world of physical IT infrastructure is an essential part of future-proofing your network. To prepare for the unknown demands of the future, your infrastructure needs to be dynamic enough to support agility and scalability. It’s not enough just to build out excess capacity.

    Excess Capacity ≠ Agility

    Improving the agility of your IT infrastructure will also improve your IT staff's ability to respond quickly and flexibly to new requirements and challenges. As our processes and technology evolve, it will be possible for technological decisions to incorporate feedback from end-users more immediately and effectively than ever before.

    That doesn't mean that making use of your new potential for agility will be easy, however. A flexible balance needs to be struck between overarching policies and localized decision-making and control. To support this, the old world of static physical IT infrastructure will be modified and expanded. In the new world of physical infrastructure, network administrators will be able to see and control every aspect of the network in real time, all the way down to physical layer connectivity. 

    Organizational silos are a factor here, too. Network managers are used to being responsible for overseeing the digital aspects of the network, but they have rarely had visibility into the physical infrastructure. Facilities managers are responsible for the establishment and maintenance of physical infrastructure, but have limited interaction with the network managers. Changing those relationships and areas of responsibility will take just as much planning and care as introducing changes to the physical network architecture.

    Another challenge is the time it takes to make changes to physical infrastructure. Depending on the scale of a request, changes can take days, months or even years. It’s no wonder that organizations turn to workarounds like SDN overlays and migration to the cloud. They need to scale and adapt connectivity and bandwidth in digital timeframes: seconds instead of hours, and hours instead of months. In the past, that simply hasn’t been possible with physical infrastructure.

    From “Plumbing” to Network Asset

    Architects and network managers are used to thinking of their current infrastructure as a fixed value, one which can be utilized in its current state but not adapted to new purposes. This “plumbing” perspective severely limits the options considered during planning, even with the introduction of virtualization and cloud resources. No one wants to invest the time and expense required to rip and replace their entire infrastructure, after all!

    As new physical layer technologies become available, that all-or-nothing mindset will need to change. There will be options for adding elements of software control to existing legacy infrastructure, as well as building out new deployments.

    Instead of “plumbing,” your infrastructure will become another network asset, its intricacies visible in real time, and controllable via software. Help your team members to embrace this shift and keep it on course by making sure that everyone understands the benefits. Your technicians are already struggling to keep up with the demand to do more with less. They will feel the value of a “digital” upgrade to your IT infrastructure even more than the C-suite.

    • Manual Documentation & Troubleshooting: Cable connections between devices are commonly tracked in spreadsheets, manually entered (or not!) when operators physically plug or unplug connectors. Changing connectivity is a time consuming and error-prone process, and troubleshooting is even more so. With the introduction of dynamic managed connectivity, tiresome manual documentation and the frustration of out-of-date connection spreadsheets will both be things of the past.
    • Overprovisioning: A common approach in legacy data center design is to “overprovision,” or build out excess capacity in the anticipation of spikes in demand. SDN and virtualization have started to replace this approach, but enterprises worried about lack of flexibility in their infrastructure still rely on racks of idle equipment, wasting power and space. With the introduction of physical layer SDN, networks can make more efficient use of their resources, reducing the need for this kind of “peak” capacity planning.
    • Rigid cooling and power: Much like overprovisioned network equipment, cooling and power capacity are planned out for peak periods, leading to wasted or unused resources during normal operation. There is also a risk of equipment failure if the new peak demand is higher than planned for. Integration of DCIM systems with SDN orchestration will lead to “just in time” cooling and provide real-time responsiveness to unforeseen situations.

    If you enjoyed this post, download the ebook today!

    Download the "Digital Transformation Road Trip" ebook!

    You can also read the original 6-part blog series online:

    Week 1 introduced the idea of the Digital Transformation Road Trip, and the question of what’s holding it up.
    Week 2
    took a look at who has a role to play in digital transformation, and how to get them on board.
    Week 3 examined the problem of approaching digital transformation as a one-time project, and what to do instead.
    Week 4 addressed the internal resistance to digital transformation that is inherent in your organizational silos, and some thoughts on how to overcome this challenge.
    Week 5 broke down another impediment to effective digital transformation – the infrastructure bottleneck.
    Week 6 wrapped it up with a look at how networks and data centers can become more agile, with a dynamic and managed physical layer. 

     

    Topics: digital transformation