Network Automation – A Mission-Critical Piece for Success

Daniel Sakimura – Data Operations Professional, Colorado Convention Center

Today’s Event and Convention networks are far different than those of the last decade.

We live and work in an era of “borderless” networks, where individuals and businesses expect to access their content from anywhere at any time.  The number of vendors, languages, protocols, and threats that network engineers have to be aware of and competent with has increased significantly.  In the convention IT business, we have seen a growing need for network services as part of the core of successful events. Networks which allow organizations to extend their office into the convention venue transparently.

End-user expectations have changed as well.  Users are no longer content to wait until tomorrow (or next week) for services to be provided.  The network is expected to “just work”, especially when the user walks into a convention venue, their office away from the office.

The network is no longer just “the network.”  It is a mission-critical piece for the success of an organization’s event.  Network services have no value if they cannot meet the client’s growing requirements.   With the rise of various cloud platforms (Software/Platform/Infrastructure as a Service) and other rapid- and flexible- deployment technologies and strategies, the network is being forced to change rapidly and flexibly.

Within the sphere of conventions and events, automation of network processes allows us to serve our customers the products they need faster and with greater consistency.  Automation of the troubleshooting process provides for a quicker and more efficient time-to-resolution.  Computers, by design, are very good at doing the same thing over and over; humans, by nature, are not.  Automating deployment and troubleshooting processes creates consistency.  Automation also allows (some would say forces) better documentation.  The programmatic nature of automation means that changes must be determined before a change is implemented, thus reducing the use of one-off, quick, “just-make-it-work” changes that cripple a network later.  In the narrowed time frame of a convention, consistent, accurate and timely delivery of services and solutions is the way we enable our client’s events to succeed.

If automation is so great, why hasn’t every network already been automated?  Automation is not easy and not for the lazy.  No network is the same as any other, and no one-size-fits-all solutions exist.  Many good frameworks exist, mostly borrowed from the DevOps movement (Ansible, Puppet, Chef, Jenkens, Salt, etc…), but they are frameworks, tools, not complete solutions.  Many network hardware companies (e.g., Cisco, Juniper, etc…) are rolling out hardware systems that can be accessed programmatically through various APIs.  However, someone still has to tell the equipment what to do, this, therefore, becomes the network engineer’s new job.

With the increasing dependency on well-functioning networks for successful events, we must begin to think broader.  It is no longer just about delivering an IP to a point in space.  Service has become about adapting to unforeseen circumstances before the client notices an impact on their operations and not just their network. Now the network engineer needs to move out of only the network and be able to work with developers, operators and administrators to deliver a product that meets or exceeds the needs of our convention clients, both now and for the future.  Maybe, instead of being “Network Engineers,” we need to become “Solution Engineers,” enabling our clients to do what they need to do without a bottleneck.  Automation provides one of the critical components to engineering successful network solutions.