The Third Industrial Revolution: Beyond 2017
In my last post, I discussed the rapidly maturing foundational elements that will help power the Third Industrial Revolution. We’re in an exciting time in IT. Everything we see and touch is smarter and more connected than ever. Machine learning is empowering the rapid evolution of various systems. Virtual and augmented reality are changing how we interact with what’s real or what we’d like to experience. Healthcare is increasingly safer and more precise, and people anywhere can learn practically anything. No doubt—there is much to look forward to in our technological future.
In this post, I’ll focus on a few areas of innovation that are under the radar to some but can have a significant impact in the future.
There’s a lot to love about building and deploying applications without having to even worry about the traditional server-oriented (physical, VM, container, etc.) deployment, management, and operational challenges. While projects such as AWS Lambda, Azure Functions, and the open-source MIT serverless project gain increasing interest, enterprises are rightfully concerned about the stickiness of provider-centric projects.
In addition, operational and security concerns including audit logging, data management, and performance management don’t just disappear. Networking and interoperability with a variety of non-serverless services is also a concern. The bottom line: there is plenty of room for innovation with regard to flexibility and heterogeneity in serverless architectures, along with solving a variety of operational challenges. This is absolutely an area to watch. In the near term, 2017 will be a big year for operational maturity in cloud-native applications.
Ephemeral Infrastructure Fabrics
Stateless, ephemeral containers are changing the way we approach application security. I have worked with an organization that builds and tears down containers on a per-transaction basis. Their reasoning is that any container compromise would be very short-lived (typically less than a few minutes or even a few seconds). Naturally, there is no technical barrier for extending the same approach to the infrastructure fabric—ensuring the infrastructure fabric does not outlive its associated application containers.
Highly dynamic, purely software-based infrastructure fabrics can and should evolve to be just as ephemeral as the applications they support. Of course, this can dramatically change our approach to security as well as completely alter how we manage application performance and resolve performance issues in the future. Think for a moment if your network and data fabrics were destroyed following the completion of a transaction. How would your current operational model handle such dynamic infrastructure?
With today’s increasingly sophisticated attacks, ephemeral services—inclusive of infrastructure—are only a matter of time. Organizations need to be experienced in operating fully software-defined stacks to be able to take advantage of ephemeral infrastructure fabrics, and network virtualization is a great place to start.
Intelligence at the Edge
Cloud computing has ushered in an era of massive consolidation, and we’re far from finished. That said, smart devices are seeing exponential growth and microservices and cloud-native application architectures are yielding applications with increasingly smaller footprints. Compute capacity everywhere in smart devices at the edge and lightweight, highly distributed applications are the perfect storm that can yield highly decentralized, distributed application stacks. The stacks will run applications where proximity to the consumer or proximity to data is paramount.
In basic terms, smart devices at the edge can act as a platform that hosts virtually isolated and highly distributed applications. Smart electrical meters installed in homes may create new business opportunities in terms of B2B services for energy companies. The compute capacity of electric vehicles could be used in a data processing grid while the vehicles are idle at night. These multi-purpose roles of smart devices can yield lower costs for consumers, with consumers opting-in to lower cost device options that simply find other tasks to do when they’re not needed by the individual consumer. This added intelligence isn’t a replacement for cloud computing but can certainly augment it. Of course, a massive application fabric, management, and security ecosystem for applications running on these devices will provide the genesis for a new wave of tech startups.
Updates Are a Feature of Software, Not Services
Increasing business agility requirements have had a profound ripple effect on IT organizations. IT’s mandate to deliver solutions with greater speed, flexibility, and security at a lower total cost of ownership (TCO) has not just impacted IT organizations but the IT services industry at large. While there will always be integration challenges unique to individual end-user organizations, software stack upgrades will be features of software and will no longer require massive professional services engagements. Enterprise software vendors that do not get on board with this trend will increasingly find themselves less relevant. The same holds true for IT services organizations with significant revenues to be disrupted by software replacing what has been traditionally labor-intensive projects. VMware is fully on board with this trend as evidenced by industry-leading support for non-disruptive OpenStack upgrades, seamless deployment, and automated infrastructure and management software updates in Cloud Foundation.
Successful IT Operations teams have to remove the phrase “We have to regain control” from their vocabularies. No Ops-centric portal can be the gateway to all innovation. Cloud computing is simply moving too fast for such an approach to be feasible. At the same time, even when brokering users to cloud services using a master corporate account, there will inevitably be times when new applications or services are provisioned out-of-band. That said, even in those cases, the security of those applications is still likely the responsibility of corporate IT.
This is where we need to be more creative and evolve how we discover and onboard new applications and services. A forced, rigid procurement model will never be a one-size-fits-all solution. We need to be smarter about discovery, and build capabilities to find new workloads and protect our sensitive information without impeding innovation. There is a new middle ground, and it doesn’t start or end with “regaining control.”
Happy New Year!
Those that know me best know that I mostly worry about what I’m not worrying about. That’s why my favorite part of my job is getting out and meeting with lots of smart people, hearing new perspectives, and learning something new. While we’re winding down 2016, I look forward to 2017 with enthusiasm and optimism. What am I not worrying about? I hope to find out soon.
Read more from VMware executives on what’s to come for technology innovation in 2017 and beyond.