Cisco’s new intent-based networking software has the IT community talking, but is this solution truly innovative, or is the company simply hoping to be associated with a new industry buzzword?
Ever since last June, when Cisco revealed its new intent-based networking software, industry players have been racing to get ahead of what its proponents call “the new era” in networking. The technology will allow network managers to automate some configuration tasks based on business intent that would otherwise require extensive manual coding, freeing up capacity for IT staff tasked with administering access to more and more connected devices at once.
Cisco CEO Chuck Robbins claims this paradigm will “redefine the network for the next 30 years,” and there’s no doubt that the ability to automatically detect malware threats in encrypted traffic will be a boon for IT professionals. But is this technology really new? Is Cisco ready to harness its power to change the industry? Or is intent-based networking less of a groundbreaking innovation than it is promotional fodder for its would-be pioneers?
The ability to perform daily networking tasks like configuration, provisioning, and troubleshooting programmatically has long been an ambition within the IT community — one that has been made far more urgent by recent developments in tech. One of the first attempts at it came as far back as 1997, when AT&T’s GeoPlex used network APIs and dynamic Java to create a sort of middleware platform.
This was the beginning of software-defined networking (SDN), a key component of intent-based networking (IBN) that allows for the adjustment of network devices through a software interface. When OpenFlow was introduced in 2011, it was quickly adopted by network administrators in tandem with SDN to push low-level rules to switches across vendors and scripting languages, effectively automating many traffic management tasks.
This arrangement has several critical limitations, however: beyond a centralized design that leaves it vulnerable to cyber attacks, OpenFlow doesn’t provide the flexibility needed in today’s IT environment. As David Lenrow wrote in 2015, when multiple SDN services use OpenFlow to push rules, “there is always a risk of conflicting changes to the system state. Attempts to examine these rules (of the form “match this header and perform this action”) and resolve such issues have been unsuccessful because at this low level of abstraction, it is impossible to decode the overall intent of the services pushing the rules.”
Intent-based networking attempts to avoid these types of conflicts by ensuring that SDN controllers need only communicate the intent of each action, rather than the protocol for achieving it. For example, instead of telling the SDN to configure X parameter using Y command on vendor Z’s CLI, applications would simply tell the controller to forward traffic from point A to point B. “Because the intent-oriented description conveys the why, rather than the how, it is possible to determine actual or apparent conflicts and seek ways to fulfill the cumulative intent of the multiple-client services,” writes Lenrow.
The IT community has made clear for some time now that there’s a real need to create this kind of scalable, flexible, portable form of networking through intent. Independent groups like the Open Networking Foundation (ONF), OpenFlow’s creator, had been researching and developing technology that could achieve precisely that for several years.
In fact, Cisco’s announcement comes just as the ONF plans to move its R&D on intent-based networking to the Managed Extensibility Framework (MEF). Given Cisco’s involvement in the ONF’s work, this has led many to suspect that the IT giant is attempting to claim intent-based networking for itself before the open-source community agrees on a set of standards that can govern its implementation.
“Whatever [Cisco is] doing doesn’t mean we don’t need a standard approach,” CTO of MEF Pascal Menezes told SDX Central. “Intent has to be implemented in a way that we all agree on.”
And while those at the highest level of the company are certainly gung-ho on the technology, it remains to be seen how the company’s huge team of engineers will react to it. “There’s a huge base of tech people who get paid a lot of money to configure Cisco equipment,” said an anonymous SDX Central source. “They don’t like this automation stuff. They learned the CLI way of doing things. If things are automated this attacks their core skill base.”
While it remains unclear whether Cisco’s play for intent-based networking will be successful, the renewed interest in the concept demonstrates the urgent need for new ways of supporting exponentially growing volumes of traffic networks must support. And as has been the case throughout the IT industry for the past few decades, solid network design will be foundational to the success of such initiatives.