The enterprise cloud market experienced a substantial amount of change in 2014. The year marked a significant increase in technical maturity for nearly all Platform as a Service (PaaS) providers and demonstrated that more enterprises are betting on PaaS as the cornerstone of their private/hybrid cloud strategy than on IaaS. I’ll be adventurous and use the experiences we’ve had at Apprenda to make some key predictions for 2015:
1. The PaaS standard will be to containerize.
Apprenda has been a container-based PaaS since its inception. Now that Docker popularized the concept and the world has vetted the viability of a container-based architecture, enterprises are moving from asking about containers as a roadmap item to asking for implementation details. While 2015 won’t mark the year of broad-based customer adoption, enterprise psychology will shift to expecting that PaaS is synonymous with use of containers.
2. Practical microservices capabilities will be favored over empty posturing.
Most customers I’ve talked to said that the microservices “advice” given out by many enterprise PaaS vendors is, at best, laughable. Many vendors have repackaged the service-oriented architecture (SOA) conversation and are delivering it under a new moniker. While doing this, these same vendors have avoided being held accountable to microservices at a feature and execution level. 2015 will require that PaaS and cloud vendors deliver practical guidance driven by core enterprise PaaS features if those vendors intend to be taken seriously by customers.
3. The Internet of Things will be a driver for implementing PaaS.
PaaS exists to support core use cases. Most PaaS implementations are there to simplify the IT model so developers can quickly build cloud-enabled applications. The Internet of Things, however, is creating new pressure to ensure that the web-service backends that power the IoT are easy to manage, highly available, and massively scalable. PaaS will prove to be the best mechanism for powering most IoT strategies.
4. Containerized OpenStack will emerge and create confusion.
There is plenty of talk of OpenStack directly supporting containers. This will interest people but adoption will be flat. Unfortunately, containerized OpenStack will create significant confusion since many have come to expect that container control and management happen at the PaaS layer. Given that PaaS is becoming the dominant form of cloud assembly, containerized IaaS will fizzle since it will directly conflict with continued growth in enterprises deploying private / hybrid PaaS regardless of whether they’ve built IaaS already.
5. PaaS buyers will dismiss solutions that are prescriptive about infrastructure.
Earlier last year, I wrote a post about how many PaaS vendors have a lot of marketing around being portable, but in reality, their products increase IT risk and drive lock-in by deliberately creating stack dependencies. We’re finding that more often than not, customers are opting to challenge vendors on infrastructure portability as early as the proof of concept (POC) phase. Customers want an enterprise PaaS that doesn’t favor one infrastructure over another and to ensure this outcome, customers are using POCs and RFPs to demand that PaaS vendors prove that their solution is portable across infrastructure solutions.
Most of what I’ve predicted for 2015 is based on my observations from real end-user data points in 2014, so I’ll hope for a successful retrospective on December 31 of this year.
Let us know if you agree, disagree, or have more trends to add to this list. Leave a comment below or respond to Sinclair on Twitter at @sschuller.