creative soul - Fotolia
I'm going to the Amazon Web Services re:Invent conference in November, but it wasn't easy to make that happen. IT departments carve out limited budget for personal development -- too often, none at all. Making the case for an enterprise network administrator to go to a cloud conference -- and some would argue the cloud conference -- is not for the faint of heart.
I'm grateful to have relatively enlightened management who trust I'm doing the right thing for the business. And make no mistake: This is the right thing for the business. Today's IT needs at least one administrator who knows not just how to keep data safe and performing well on a hybrid network of on-premises and cloud connections, but how to make off-site services deliver on their enormous promise. If you play your cards right, you might not only gain skills that save the day at work, but gain a little Cloud Jockey swagger as well.
The smaller number of networking conferences versus application and data center shows is actually somewhat of a blessing. The major events, such as Interop and Cisco Live, concentrate networking vendors and useful session content into larger events, colocating broad-topic, best-practice, how-to and narrow specialty expert content. It's easier for network event attendees to mix and match dense, four-day schedules aligned to their interests and business needs. Hit Interop or a Cisco Live, and you'll learn something new every year.
For application administrators, however, it's more difficult to know what to attend and try to justify to the business, with so many separate specialty conferences for virtualization, storage and each major application group. The conferences tend to be smaller, and it can be more of a challenge to build a meaningful, full schedule. So, a loud-and-cloud event like re:Invent must be an even more exotic, unjustifiable distraction for network engineers, right? Wrong.
Why hybrid network training is a must-have
Every time an IT manager pushes yet another cloud or software-as-a-service (SaaS) project, a network administrator somewhere experiences a flash of heartburn. We've done so much to nearly bulletproof our internet links that our immediate reaction to management statements like, "We're moving all our sales tools to SaaS," is now something other than, "Aw, hell no." We've engineered redundant fiber to opposite physical sides of our core, mitigating even random backhoe misadventures. We identify service issues not just inside our networks, but through complex multihomed internet paths we didn't even know about -- all the way to service endpoints inside our SaaS providers.
The downside is even if your business isn't e-commerce-based, more likely than not, there's a revenue hit every minute you can't reach critical off-site applications. You've probably got at least some AWS or Microsoft Azure machine instances running in the cloud -- and even more likely, Amazon Virtual Private Cloud (VPC), with dozens or hundreds of applications flying in formation with your on-premises systems via connected networks. Hybrid IT is today's reality.
But, as the network administrator, are you as comfortable configuring and troubleshooting internal AWS Direct Connect as you are your gateway router? What's your automated failover policy for VPN backups? What's your access policy approach for securing public versus private interfaces? How do you tune your physical router configurations to optimize VPC connections? What's your approach to VPC IP subnet management? What's your holistic monitoring approach for troubleshooting service-related help desk tickets from end users? Which works best for configuration, Chef or Puppet? Ah, that stuff.
There's a conference for that
These are the questions that keep us up at night. It's so easy to get started with a little trial and error in the network tabs of AWS or Azure consoles. To speed adoption, these cloud providers conceal critical long-term network operation concerns. Right now, if your physical network was flashed back to Dark Age defaults by a global electromagnetic pulse, you could rebuild it. You have your configuration backups. You could rerun discovery, and in the worst case scenario, walk the campus with a laptop and perform the healing, laying on of hands. But if someone accidentally breaks your VPC endpoints, connections, endpoint router configurations, internal access policy configurations or worse, you will have to be just as handy and heroic as you are with your on-premises network.
Your business would certainly benefit from you learning these skills from experts. With hybrid network training, your career will benefit from staying ahead of the transition to hybrid networking, especially certification. Will there be a lot of application-focused tech at re:Invent? Sure. But you'll notice under almost everything, there's as much concern for the network, even network as a service, as there is for on-premises networks. Cloud is as much a network as it is compute or storage. Why wouldn't network administrators want to be experts on the whole network they're responsible for?
So, the next text time an application developer anxiously asks how you're going to assure high-availability, low-latency, durable, secure network services for their crazy new internet of things application that has to go live in a week, crack your reassuring IT smile and say, "No problem."
"We're already monitoring the VPC, network connections and end-user quality of experience. I'll help you configure a little Lambda code to export your AWS VPC flow logs as Internet Protocol Flow Information Export, and you'll be all set."
Your manager will smile, too.
Find out what building a cloud network means
Amazon VPC is an essential piece in the hybrid cloud network
Network component crucial for cloud success
SDN is key to properly positioned cloud services
Is formal SDN certification and training worth it? Depends