In this ESG360 Video, ESG's Bob Laliberte and Mark Peters discuss current issues and trends in the world of Networking.
Read the related ESG Blog: Talking Networking, Part 2
Announcer: The following is an ESG 360 video.
Mark: I've been spending time recently talking to some of my specifically focused colleagues about the areas that they cover and the impact that that has on IT and outcomes as a whole. Last time I spoke to Bob, we concentrated on the fact that networking is so complex and integrated and important these days to actually delivering all these cool apps that we all like to enjoy. Whether that's a commercial at work or social. I'd just like to come back down and be a little more prosaic, if you like, this time and ask you to talk about some of the more interesting networking technologies that are either on the horizon or, indeed, beginning to make their foray into the market at the moment.
Bob: Sure. That would be great. I think in the past the legacy networking environment has been fairly static. People did things a certain way. Maybe we were going from three tiers to two tiers, but a lot of wired connectivity and so forth. Everything was manual. When you wanted to make a change you had to go and touch every device in order to do that, and I think what we're seeing now is a recognition that in order for the networking component of the overall IT infrastructure to be an enabler for transformation, whether it be the digital transformation, etc. Accelerated business processes.
It's not going to be able to retain or remain a manual effort. So what we're seeing is a lot of automation coming into networking. We're seeing a lot of AI and ML come into networking. In order for organizations to have basically a networking infrastructure that is highly programmable and quickly being able to scale and to respond to the needs of the business. So a lot of companies have different initiatives around that. Some of the larger ones refer to this as intent based networking and the entire focus is how do we align the business goals to the network?
Mark: So intent, is that from the vendor's perspective or is it...what I'm really getting at is from the end user, they know what outcome they want. Do they use the machine learning and the artificial intelligence to get that done or just to inform them?
Bob: Yeah, so good question. It's really about the understanding the intent might be I want to ensure that none of the guests can get onto my corporate network. And then in the old days it would be okay, I need to go out and set up a separate subnet and I need to touch every device and let it know what's going on. So the idea behind this would be simplified to say, "I need to create a completely independent and separate guest network." And then what would happen based off that intent, it would automatically create that sectioned off guest network without having to go to each and every device. So the software would go to each and every device, they would look at its state, they would configure it, they would get it set up. So what you're looking at is the ability to rapidly respond to a business change without having to go through a lot of the manually intensive steps that it would have taken before.
Mark: Because you can change the policies around that or the capacity of that particular subnetwork, if you like...
Mark: ...with a few swipes or a few clicks?
Bob: Right. You would build the policies...
Mark: Well hopefully a few voice commands these days.
Bob: Well eventually, yeah. Eventually that's the point, right? "Alexa, please configure my guest network for me." We're probably on our way there, not quite there yet, but that's the idea behind it is that instead of saying, "Okay, I'm going to send out and do CLI commands to every device that's in this network to ensure and then hope I got everything right and didn't mistype anything" that you would have a policy based back-end that, based on those policies that you want to have for those different subnets, it would then go and create that for you. So...
Mark: And a bit of a philosophical question to wrap this segment, would we see networking as more complex and capable or would we actually see it actually disappearing and just being more compliant? Would we be aware of it or do we hopefully not see it at all?
Bob: So the ultimate incarnation of this would be you would not see it. Right? You would add a level of sophistication to what is a very complex set of changes and that would all be abstracted and it would become much easier. You would get to the point where you could say, "Alexa, please do this" and then based on the policies you've established, it would roll out those changes knowing that everything would be set up and running. The best part is, also, as part of the machine learning that's being added not only would it roll out initially, but as we all know environments are not static. They change from day to day. It would also leverage machine learning to benchmark how that should be performing and be able to do a lot of self-healing adjustments to ensure that those business service levels are continuously met on an ongoing basis without potentially having to go back and do any manual changes to it.
Mark: Well I'm going to summarize with an awful, sort of, pun but it makes sense to me, which is, if you like, it's putting the working back into the network. Because a network is a very static thing. It's the plumbing and what you've described is something that's very dynamic and genuinely useful as we create this bigger integrated IT world.