A funny thing happened on the way to making enterprise IT simpler.
All the shiny things (cloud! AI! automation! IoT!) that were supposed to make IT run faster and require less hand-holding have actually added more layers of complexity while extending the boundaries of what must be monitored, analyzed, serviced and secured.
“There’s more and more stuff out there. There’s a lot of things we’ve always managed well in traditional IT. But it’s all the non-traditional IT — how do we manage that?” Gartner VP David Cappuccio wondered aloud at the firm’s recent Gartner IT Symposium in Toronto.
Cappuccio and fellow Gartner VP Thomas Bittman broke down some of that new complexity as they ran through the hottest trends in enterprise IT and cloud computing, respectively. Fortunately, they also offered tips for managing this growing complexity.
1. Serverless computing
As defined by another Gartner analyst, Ross Winser, serverless computing “is an emerging software architecture pattern that promises to eliminate the need for infrastructure provisioning and management. It enables code residing as functions to be run without the need for the user to explicitly provision or manage infrastructure.”
Although it doesn’t replace VMs or containers, Winser noted it “can support requirements for utility logic, unpredictable demand and event-driven requirements.”
At the symposium, Cappuccio said 10 per cent of IT organizations are already using serverless computing, and Gartner predicts it will become mainstream by 2022.
2. Network evolution
“The network is growing so fast, with so many services being deployed … (and) they all require network security and network engineers to pull it all together,” said Cappuccio.
As examples, he cited two Gartner predictions:
- the number of large enterprises using intent-based networks will skyrocket from less than 50 to more than 1,000 within the next year
- by 2022, the number of enterprises deploying automation in the access layer will jump by 50 per cent
According to Cappuccio, although the network keeps getting more complex and line of business expectations continue to rise, budgets remain flat, creating a crunch he dubbed “the network innovation gap.”
3. Death of the enterprise data centre
In a blog post last year, Cappuccio famously declared, “the data centre is dead.” In Toronto, he softened that just a bit, but said the traditional data centre is definitely on the endangered list.
“It’s going to take time but the (traditional data centre) is going away,” he said. “The trend now is cloud-first, on-premises last.”
By opting for as-a-service and moving workloads into co-location, Cappuccio said enterprises are starting to think: “I can use their network rather than mine. It makes me more agile and at the same time frees up my network team to focus on other things.”
In his own presentation later that same day, Bittman echoed Cappuccio’s data centre prediction by declaring “private cloud is dead.” Bittman said less than five per cent of enterprise workloads will run in true on-premise private clouds by 2020.
4. Edge computing
If you’re a business with customers scattered across the globe, or a retailer worried about losing online shoppers due to slow page loading, “network latency becomes the biggest issue — milliseconds count,” said Cappuccio. So enterprises are moving their computing to the edge where “availability is better and uptime is better.”
Bittman added that the explosive growth of connected things and more immersive human/machine interfaces (like IoT, AR and VR) also creates “a cultural and generational expectation of speed that simply can’t be met” without edge computing.
5. Digital diversity management
One of Cappuccio’s slide decks said it best: “Extreme choice is complexity disguised.”
IT isn’t just worrying about traditional routers, IP addresses, software and code anymore, Cappuccio said, but IoT, sensor batteries, firmware, physical device location management, asset placement and more.
“As we go towards this new world, all this stuff is going to be IT’s responsibility,” he said. “All these things are being driven by edge (computing) and we have to manage it all. We’ve got to figure out how to see what’s out there, and we don’t have total control of it.”
Fortunately, there are a few things you can do to manage this IT complexity, according to Cappuccio and Bittman:
Cloud: If you’re still measuring your cloud’s success by cost savings and consolidation, focus more on your cloud’s ability to improve things and deliver more, said Bittman.
Automation: “Think ‘smart automation.’ Automate tasks to free up your staff,” said Cappuccio.
Networks: Intent-based networks are going to be huge, so “at least get someone in your engineering team to understand how it will help you and impact you,” said Cappuccio.
Talent: Since the cost of hiring new talent is sky high, look at your existing staff and get them to start thinking about things differently. “Seek out adaptable collaborators” with broader experience versus narrow skills, and “essentially give them something to learn,” said Cappuccio.
Partners: Seeking strategic “partners and partner skills” is key to extending your IT infrastructure globally at the lowest cost and in the quickest, most manageable way, said Cappuccio. Good partnerships also reduce your complexity and responsibility considerably.