Myth # 4 of the Good-Enough Network: “Just Look for Standards”

Editor’s Note: This is a post by Michael Rau, Vice President, CTO, for the Borderless Network Architecture Cisco.

Just 50 years ago, personal computers, corporate networks, and email didn’t exist. So sending a note meant getting out your trusty pen and paper, writing a letter, then dropping it in a mailbox and hoping it would arrive within a couple of weeks (hard to believe, right Gen Y’s?). Today, it takes mere seconds for an email to get to a recipient.

New innovations and inventions in the technology industry mean the flow of information has changed the method and speed at which we communicate. And the standards governing the technology industry help ensure there is security, interoperability, and a framework in place. As we innovate, old standards evolve and new ones are created. Imagine if we used the post office standards from 1890 to govern the way email is sent. If that were the case, we’d probably be putting postage stamps on our email messages.

Standards vs. Innovation

Cisco has a deep respect for industry standards and participates in many standards bodies. As we’ve learned, vendors interpret and deploy standards differently in their equipment. These differences may result in integration challenges. While industry standards are extremely important, relying only on existing standards as you plan for future technology needs is misguided.

When companies lock themselves into standards-based networks, they miss out on a higher-level of service innovation and occasionally underestimate the integration cost involved in making the components of a standards-based system work together. Yes, standards should be used, but businesses looking for a competitive edge need to look for solutions that are also innovative.

Businesses are always looking for innovative ways to collaborate with their end customers, ways to better manage their infrastructures, and ways to reduce complexity. This often means investing in next-generation technology that may not yet be standards-based.

The good-enough network we’ve been talking about in the previous myths, the ones that are simply cobbled together with only cost of the parts in mind, advocates being “standards based.” This approach often means that if a customer buys industry-standard servers, storage, and networking technology, they will save money. The theory being that the network will be easy to set up and everything will work together because it’s worked in thousands of other businesses. While good in theory, the interoperability between different vendor implementations of the same standard and the lack of pre-standard innovations don’t necessarily make for a scalable, reliable, and feature rich network to support the needs of the business.  What is really needed is robust standards and innovations in the network that ultimately become the standards of tomorrow.

Setting the Standard: Case Studies

Let’s look at a few examples of turning innovation into standards. Let’s first consider the Cisco Discovery Protocol (CDP). Cisco developed CDP years ago as a way for the network to discover the device being plugged into it and to apply the appropriate configuration. Medianet was the next advance in CDP. And today, Link Layer Discovery Protocol (LLDP) is an industry standard based, in part, on Cisco’s innovation.

For the virtual data center, Cisco is developing innovations like Overlay Transport Virtualization (OTV), which allows a customer to combine two or more separate data centers into one virtual data center. Cisco also just introduced the Locator/ID Separation Protocol (LISP), which can move a virtual machine carrying workloads between data centers without having to change the address of the virtual machine. These innovations avoid the complexity of managing protocols such as Multiprotocol Label Switching (MPLS) and reduce other administrative tasks; and this, in turn, leads to quicker time to deployment–all of which saves money.  What is interesting about LISP is amount of work that Cisco is simultaneously doing to support LISP standardization efforts through the IETF.  LISP, like MPLS prior to it, is another example of Cisco developing innovations that can be deployed today while the standards effort is driven.

In addition, Cisco has fundamentally changed the way networks are designed with Virtual Switching System (VSS) technology, which satisfies three major demands of networks for high availability, better capacity utilization, and simplicity. VSS combines two Cisco Catalyst 6500 Series Switches into one logical switch, with the immediate benefit of reducing the number of switches that have to be managed by half and doubling the redundancy. In addition, with VSS, the access layer device doesn’t see the merged switches as two separate data paths, but as one giant pipe. The cost savings mean reduced device management expenses, improved uptime, and greater utilization.

Yes, we do think standards are important. Whatever future technology a customer chooses to invest in, they should select a vendor who is committed to existing standards. But building next-generation network means also taking into account innovations that may not yet be based on a standard, but will increase speed, efficiency, and improve business operations overall. Innovations that will bear fruit both now and down the road.

Just think, if we didn’t ever look past existing standards, we might still be sending letters the way our grandparents did. And you’d be reading this article on paper instead of online.

Four myths down… three to go. Stay tuned for myth number five next week.

What are some of the “good enough” myths that you’ve been hearing in the industry?

Note:  The seven myths are outlined in a recent white paper from Cisco: Debunking the Myth of the Good Enough Network.