pointer

Tag Archives: sdn

The Software-Defined Data Center and Software-Defined Networking: What Does It All Mean?

Posted on by .

    Software-Defined Networking (SDN) has become the center of heated discussions in the IT community, just like Big Data and The Cloud. This year, another term is being added to the list, the Software-Defined Data Center (SDD). As with any new trend or buzz word there is the question on whether these are valid, and if they will truly live up to their perceived potential.

    This blog will help define these technologies, what stages they are in now and what segments of the IT department these technologies might help or hinder in the interim.

    What is The Cloud?
    While there’s not a lot of debate on what cloud technology is, starting with this term will help the discussion by introducing some key ideas.  While the canonical document comes from NIST, there is a much easier definition from Cloud Camp founder Dave Nielsen. In his view, cloud is OSSM (pronounced “awesome”):

    • On-demand: the server is already setup and ready to be deployed
    • Self-service: customer chooses what they want, when they want it
    • Scalable: customer can choose how much they want and ramp up if necessary
    • Measureable: there’s metering/reporting so you know you are getting what you pay for

    The OSSM nature of cloud is driven by automation, controlled through orchestration. Customers use a cloud control panel portal, or an API, to monitor and configure their cloud instances at-will.

    What is Software-Defined Networking?
    SDN is a movement to apply automation and orchestration to networking equipment, in a manner similar to what cloud computing has done for virtual servers. The current common practice for configuring network devices requires a skilled engineer to configure each firewall, router, or switch separately when making a network change. In the words of Packet Pushers Podcast host Ethan Banks,

    “To add a new VLAN and properly advertise it throughout the LAN/WAN, there are many steps required[…] I need to touch core switches, access switches, firewalls, and WAN routers[…] what should be a routine task is too darn complicated and mistake-prone.” (from the section titled “So, What Have We Got?”)

    A single misconfiguration may lead to a cascade that takes down vital services – like GoDaddy on September 10, or Twitter in June. Therefore, it makes good sense to use orchestration to push configuration changes where they need to go, automating the process to reduce the amount of work and the number of errors.

    As more companies are acquired and more companies start to use software-defined networking and OpenFlow, there is no doubt that this has huge potential. However, while centralized control of a network sounds good in theory, migration may require replacing a lot of hardware and the task of creating network-wide policies to emulate the current configuration would require a lot of effort.

    What is the Software-Defined Data Center?
    SDD combines techniques of cloud computing and SDN into a manageable scope. Datacenters are often the first places that new technology is deployed in production, such as SAN and lossless Ethernet, so adding enhanced orchestration is a natural fit to increase efficiency and reduce downtime. Virtualization architect Stuart Radnidge put forth an excellent vision of the power of SDD in his blog post “On the Software Defined Data Center.”

    The essential idea of the software-defined data center is that specialized software will replace specialized hardware throughout the data center, reducing the tedious configuration work on a per-server and per-network-device basis. Derrick Harris of GigaOM describes the conversation he had in May with Steve Herrod, the CTO of VMware, about his vision and why this technology is important. At VMworld, the company came out with vCloudSuite 5.1, which really catapulted the phrase SDD onto the IT buzzword list.

    SDD also has the potential to create virtualized converged overlays onto datacenter equipment, allowing servers to interoperate seamlessly with storage and networking.  Additionally, policy enforcement will allow logical separation of co-located equipment into separate virtual datacenters, reducing the concerns of multitenancy.

    It will be interesting to see how these technologies evolve this year and into next. As acquisitions continue to be made and new products come out, each of these technologies will face challenges to live up to the hype.

      Trends Affecting Network Engineers Today – From Software Designed Networks to Mobile

      Posted on by .

        The IT world is currently in the throes of a huge shift — a seismic shift that the industry historically experiences every five to ten years. Today, the entrance of technologies like mobile, software defined networks, virtualization, and cloud computing have changed the landscape for both the consumer and the enterprise.

        Change is of course inevitable, and welcome, and while all of these technologies have either great potential or are already helping to fuel better productivity within IT, there are many unanticipated challenges cropping up. Below we take a look at some of the challenges these top trends are introducing, and how to adjust so your organization can get the most from these new technologies.

        Software Defined Networks and OpenFlow
        Software defined networks (SDN) and OpenFlow have been touted as enabling technologies that will help decrease the complexities of cloud and virtualization. SDN defines the overall technology, while OpenFlow is a specific example of an SDN, and was created as a programmable network protocol to help manage and direct traffic among switches from an assortment of vendors. Ideally this would provide centralized control and easier network management of potentially cheaper switches without the single-vendor lock-in.

        However, these technologies present potential challenges for network engineers. While the promise of centralized network control sounds good in theory, the migration to OpenFlow requires creating new network-wide policies. It’s likely that we’ll hear about large “failed” OpenFlow deployments, where the amount of effort overwhelms the projected ROI. The challenge for OpenFlow now is to live up to the hype: deliver demonstrable performance improvements without requiring a forklift upgrade of the network core. While it’s exciting that OpenFlow has lots of potential, if it’s too hard to deploy, it will never truly leave the research environment where it was born.

        For more details on the history and use cases of SDN and OpenFlow, check out our blog “Software-Defined Networking and OpenFlow to Infinity and Beyond.”

        Virtualized Networks
        More and more companies are turning to virtualized environments to streamline application deployment, to simplify IT operations and to allow IT organizations to respond faster to changing business demands. With decreasing prices and an increase in administrative tools that make management easier, virtualization is now being adopted even by smaller mid-market organizations.

        But virtualization creates “blind spots” in your network, areas where application traffic cannot be properly monitored with traditional techniques, opening the network up to undetected application performance problems. In a traditional server environment, you would normally span a switch port from a physical Ethernet switch or router and the data would stream across into a network/application performance analysis appliance, providing complete visibility. But in the case of a virtual environment, data comes back through a virtual adapter without actually hitting a physical switch. This creates a blind spot in your appliance and the communication between virtualized applications on the same server is never seen.

        In order to combat this blind spot and successfully perform network analysis in a virtual environment, you must plan ahead. Although there is no big difference in network analysis techniques in a virtual environment, there is in the implementation. Instead of capturing data at the physical layer, you must be prepared with a solution that can collect data at the level of the virtual switches.

        Mobile and Wireless Networks
        In today’s digital age, wireless networks are essential to both businesses and consumers. However, maintaining strong performance and security of wireless networks can be difficult — especially in the era of BYOD (Bring Your Own Device). And keeping up with the pace of technology can also be challenging, with 802.11ac and 802.11ad right around the corner.

        The introduction of wireless-enabled smart phones and tablets has ushered in new challenges for wireless network management, most importantly in the areas of security and performance. Now on top of dealing with the authorized workstations, network admins must account for and secure a whole new set of devices, which are not within their direct control. And when it comes to performance, not only do more devices make for a more congested wireless network, but a powered on, inactive smart phone that is not connected causes at least ten times as much damage to your Wi-Fi network as the same phone when it is connected (see http://www.sniffwifi.com/2012/04/phones-on-wlan.html for all the details).

        You need a full-featured wireless network analysis solution in place and monitoring your network 24×7, searching for unauthorized devices and analyzing overall network conditions, like excessive probe requests/responses that can drag down your overall aggregate WLAN throughput. You also need a solution that will future-proof your investment as 802.11ac and 802.11ad begin to take hold.

        Change is always a mix of good and bad, but with these new technologies come a plethora of new opportunities. In order to stay ahead of the curve it’s important to know how and when to adapt, as well as which tools will help you get there.