YOU ARE AT:OpinionReader Forum: Cisco’s white box software should extend into the cloud

Reader Forum: Cisco’s white box software should extend into the cloud

The very software technology that’s threatening the incumbents is also providing a solution that may ensure their survival.

Editor’s Note: In an attempt to broaden our interaction with our readers we have created this Reader Forum for those with something meaningful to say to the wireless industry. We want to keep this as open as possible, but we maintain some editorial control to keep it free of commercials or attacks. Please send along submissions for this section to our editors at: [email protected].

The recent reports that Cisco will package its software as an operating system point to how software is helping evolve legacy network infrastructure.

According to unconfirmed reports from The Information and others, Cisco is planning on separating out the software that is currently bundled with its networking devices. The software will be packaged as a standalone operating system, codenamed “Lindt,” capable of running on general purpose processors and white-boxed servers instead of Cisco’s purpose-built silicon and proprietary hardware.

In the near term, the move appears to be Cisco’s answer to software-only routing and switching competitors, and the days when switching and routing necessarily meant “Cisco,” are likely numbered.

“Cisco’s switching share has declined from around 70% in 2010 to the mid-50s today as Arista and white box alternatives have won share in the data center,” said James Crawshaw, a senior analyst with the Heavy Reading market research group.

Decoupling the routing and switching code from the underlying hardware is effectively an embracement of software-defined networking. It will afford Cisco far more agility in how the company evolves, packages and delivers its networking feature set. At the same time, shifting to selling software could undermine Cisco’s revenue stream from selling, maintaining and upgrading the underlying appliance hardware. In that light it’s no surprise that Cisco has not yet confirmed Lindt development.

Cisco needs to join the SDN market, but only on its terms so as to protect its market position. It is very much the same kind of dance that carriers are going through as they embrace software-defined wide area network services as a means to protect their position in multiprotocol label switching services.

In both cases, Cisco and the carriers are hard pressed to differentiate their legacy solutions from new competitors. Cisco is fighting third parties that have proven themselves adept at creating scalable, robust routing and switching code. Meanwhile, carriers are finding SD-WANs that have shown how direct internet access can often be used to replace MPLS.

Ironically, the very technology that’s threatening the incumbents is also providing a solution that may ensure their survival. By embracing SDN, Cisco hopes to extend its legacy infrastructure software. By delivering SD-WAN services, carriers want to extend the lifespan of its MPLS services.

Lindt meets the cloud?
The comparison to SD-WAN services is telling in other ways. The use of SDN principles suggests that the Lindt operating system would run on premises in the data center, for example, controlling the resident switches and infrastructure. Such an architecture makes sense, particularly given the latency requirements of intra data center communications.

However, running a controller in small offices may be unrealistic, although there’s a major need for those capabilities. Cisco’s infrastructure business in the branch offices is also under fire whether from competitors targeting low-end routers and switches or SD-WAN vendors looking to own the branch office edge. In Cato Networks’ recent survey of more than 700 networking, security and IT professionals and executives, we found that 39% (and more than 50% of CIOs) plan to eliminate hardware appliances in 2017, which may point to some of the rationale behind Cisco’s decision.

Cisco could address distributed offices far more effectively by shifting Lindt to the cloud. The white box equipment would connect back to the cloud not to a local controller. As a multitenant service there are no servers to deploy on premise or virtual appliances to scale in the cloud.

The idea isn’t so far fetched. For example, the company has already shown its willingness to deliver cloud services with the introduction of Cisco Umbrella earlier this year. A Lindt service would do for networking what Cisco Umbrella does for security.

A cloud-based service would address other strategic challenges that companies face as well existing challenges with any appliance-based solution. Today’s mobile-centric workforce needs networking and security services, which cannot be easily and effectively delivered through server and appliance. The same is often true with the internet of things. New kinds of IoT endpoints are being deployed in unmanned sites or other difficult-to-reach locations. Expecting to use appliances with their penchant for patching, maintenance and upgrades for these use cases is unrealistic.

As a cloud service, deployment and delivery of Lindt, even to remote or unmanned locations, would become possible and be far faster than appliance infrastructure. It’s just a matter of connecting to the service, assuming of course that the service provider has already build out or can rapidly expand its network footprint.

Cloud services also scale far better than appliances. The unlimited compute cycles and storage resources of the cloud eliminates the forced upgrades that come with appliances. Cisco customers can run relatively “dumb” consumer premise equipment with most compute-intensive functionality of Lindt running in the cloud.

Visibility is also enhanced as cloud providers gain almost immediate insight across their customer base. Once appliances are installed, appliance vendors lose any real-time insight into customer traffic. With deeper visibility, cloud providers are able to identify emerging trends and threats far faster than most individual organizations.

Product evolution also occurs far faster in the cloud than rolling out new features on appliances. With appliances customers need to test and roll out patches and upgrades, creating lag between a vendor’s most current software and the software appliances running in the field. By contrast, with the cloud Cisco would be able to make new networking features available instantly to all service customers, independent of the customer adoption lifecycle.

Multitenancy: a must for the cloud
But, to achieve these benefits, multitenancy will be essential to Lindt. Running virtual images of networking appliances in the cloud bring few benefits as service providers must still maintain, scale and manage each customer instance. These instances may share infrastructure in theory, but in practice that poses risks. When new features are enabled in one instance or traffic volumes surge, compute requirements leap and could impact the other instances sharing the infrastructure. Multitenancy avoids those problems by providing a single resource that is scaled, updated and maintained for all users.

The shift to software is the shift towards a more agile, more efficient infrastructure. The cloud is the natural place for delivering infrastructure software at a global scale and we’ve seen it transform practically every aspect of IT. Applications, platforms, infrastructure – all have migrated to the cloud. Now it’s time for networking and security to join them.

ABOUT AUTHOR