Here's the crazy thing about looking ahead at this particular point in time: Quite a bit of enterprise tech that got traction this year was barely on the map last year. How much of it has legs? A whole lot of it, I'm betting, which indicates a pace of change that's much quicker than usual.
1. The triumph of the public cloud. This year, IaaS and PaaS merged, making it easier to build, test, and deploy applications in the public cloud. All the major public clouds now provide both in some sort of integrated fashion, with Amazon Web Services now offering multiple PaaS options.
Meanwhile, the private cloud has stalled due to the cost and complexity of enterprises deploying and maintaining the whole stack in-house. Cloud innovation is where the action is in enterprise tech, so I have to wonder whether any business can keep up with the velocity of technology change right now. Regulatory hurdles and sunk cost aside, why not simply move to the public cloud? After all, it's the public cloud provider's job to stay on top of every new advance -- enterprises, not so much. Sure, migration will take time, but companies like GE are already declaring they're all-in.
2. Container madness. Docker, the hottest open source project on the planet, enables you to package applications so that they'll run in the containers built into the Linux kernel. Why is it such a big deal? Because it means genuine application portability -- using lightweight packages instead of a full VMs. Docker the company is working with Microsoft to create Docker-driven containers on Windows as well. Most people talk about using Docker to move apps from dev to test to production, but I'm convinced Docker will also be used to move production apps among clouds.
Moving a single packaged app from container to container is easy, but complex apps involving multiple containers are much harder. Here's where the Docker ecosystem comes in: Docker management and orchestration tools help you assemble and move complex apps with many components running in many containers. Top projects include Kubernetes, Mesos, and StackEngine; Google Cloud Platform and Amazon Web Services added their own container management systems last month.
3. Microservices architecture. In contemporary Web and mobile app dev, developers tend to build applications from services rather than writing everything from scratch. Typically, these services are microservices -- single-purpose, API-accessible apps crafted to become the building blocks of larger applications. Docker has accelerated the microservices trend by providing a handy way to package and deploy them.
If you recall the SOA trend of a decade ago, microservices architecture may sound familiar. The main difference is that microservices architecture looks at services from a developer's perspective rather than an enterprise architect's perspective, so the services are finer-grained. Also, communication among services is simpler: JSON replaces XML and REST replaces SOAP, and heavy-duty middleware is not part of the deal.
4. Liquid computing. InfoWorld executive editor Galen Gruman coined the "liquid computing" phrase to describe the effect of ad hoc networking among personal devices, where you can save state as you move from smartphone to laptop to tablet to desktop. For example, if you're in a meeting and adjusting a presentation on your tablet, when you get back to your office, you'll find that presentation served up on your desktop front and center. The first taste of this is the Handoff feature in OS X Yosemite and iOS 8, but Microsoft and Google are working on similar schemes for their device ecosystems; Samsung also recently announced its own version.
5. Multicloud management. Clouds tend to be big, complicated platforms. The more you build on any platform, the more you become dependent on its unique features -- and in the case of a public cloud, the more you're locking yourself into a platform owned and operated by someone else. Few large enterprises want to put all their eggs in one basket, which is where multicloud management comes in.
Tools to manage deployments across multiple clouds emerged a while ago and are gaining more traction. CliQr, a multicloud management startup backed by Google Ventures, claims to be able to determine dynamically which clouds should run which workloads. But a number of others, notably RightScale, enable you to manage and optimize resources and costs across clouds.
6. Endpoint security innovation. Enterprise security will remain in a desperate state as long as systems remain unpatched and untrained users continue to accidentally download malware. Nonetheless, I've been impressed with several new security solutions that emerged this year. The first, Tanium, applies innovative search techniques to interrogating endpoints across the enterprise. Tanium can obtain a near-real-time view of hundreds of thousands of endpoints to detect anomalies and determine which software lacks the latest patches -- and roll it all into a dashboard view.
Interesting solutions for mobile have appeared as well -- and not just fingerprint reading. Several Bluetooth LE proximity solutions enable you to use your smartphone as a security key or pair a physical token with a mobile device for proximity-based authentication. More recently, Android 5.0 Lollipop introduced "trusted places," which uses location to remove password or pincode gates when you're in a zone you feel secure about, such as your home or office. Convenience is an important factor, because it increases the likelihood users will practice good security.
7. Machine learning. This is pretty much the new name for artificial intelligence. On the one hand, it's important not to overpromise the near-term potential of machine learning. On the other, it's essential in making sense of big data, and open source projects like Mahout and Spark/MLlib are easing the path. As James Kobielus noted earlier this year, machine learning is so pervasive, we can often assume its presence in big data applications. IBM is mainstreaming that idea by opening Watson APIs, while startups like the Andreessen-backed Adatao are applying today's abundant compute horsepower to revive neural net algorithms.
8. The return of devops. This mashup of "development" and "operations" is really about increasing the efficiency of operations in order to make agile development real. The devops trend first emerged five years ago, but vendors have revived it, grandfathering in application lifecycle management, automated testing tools, database virtualization, release automation, configuration management, application performance monitoring, platform as a service, and related technologies.
In some circles, devops is considered a way to give developers ongoing responsibility for applications in production, but that doesn't scale. It's probably best to think of devops as shorthand for the most modern and efficient way to configure dev and test environments, which must scale up to meet today's almost universal business need for more and better applications.
9. The end of network switches. No, we won't see network switches disappear in 2015. But virtual network devices, software-defined networking, and the abundant horsepower of servers are leading to a major rethink of the data center network. The long-term prospect of the network being reduced to "the wires between the servers" is becoming more real.
Cumulus Linux brings the network control plane to industry standard hardware, and within reach of today's server orchestration tools, while preserving wire-speed network operations. A recent OpenFlow project released by InfoBlox called LINCX this year shows the potential power of a completely software-programmable network. Meanwhile, NFV (network function virtualization) -- leveraging server virtualization and data center orchestration to deliver load balancing, firewalling, WAN acceleration, and other network functions as a service -- is all the rage among service providers and cloud platforms like OpenStack.
The open source imperative
A common thread runs through most of these nine trends: Open source is leading the way in technology development. It's become the vehicle of choice for startups to gain traction, as customers -- mainly developers within companies -- take new technologies for a spin, provide feedback, and eventually put them into production. Meanwhile, other developers see what's hot and start building an ecosystem around a core project, as has occurred with Docker, Hadoop, OpenStack, and others.
The simple model of open source development -- collaborative, self-organized, and distributed -- even appears to be having an impact on enterprise app dev. That trend will take years to unfold, although some companies are now experimenting with it.
Over time, it's clear that IT spending will shift from buying and maintaining hardware and software to subscribing to cloud services -- as well as investing in developer talent and tools. Creating a high volume of quality applications that differentiate a business is the most important activity in enterprise tech. Those who adopt the best processes and technologies to accomplish that goal will win.