A flurry of news reports emerged this month stating that PaaS is dead. It started with a research report by 451 Research that noted that services that offer cloud based application development environment to customers will cease to exist independently since this is already being brought into the fold of IaaS and SaaS products. For instance, SalesForce; one of the leading SaaS businesses in the world recently launched a business division called Force.com to offer their customers a development environment on the cloud. Similar services have been launched by companies like VMWare and CenturyLink that have traditionally been IaaS vendors.
Does this mean PaaS is dying? We have always had Is the concept of “platform of a service” becoming a simple feature offered by IaaS vendors? There have always been a debate on the various cloud computing products and solutions along with those provided as managed services by third parties. According to Krishnan Subramanian, the director of OpenShift strategy at Red Hat, one reason why PaaS seems to be headed this way is because the earlier days of PaaS were clouded by hype and buzz - now that this has died down, the service has now attained a level of maturity where there two distinct flavors of PaaS has emerged - one that is characterized by service orientation and another by container orchestration.The service oriented PaaS strategy is exhibited by the Google App Engine where the service resembles a SaaS architecture while the container orchestration, exhibited by Linux containers like Docker.io is a perfect example of PaaS resembling the VM-based approach of IaaS. Subramanian notes that the distainct flavors of PaaS needs to be recognized and nurtured in order for the community to derive greater benefit from what PaaS has to offer.
But according to Barb Darrow from GigaOm, a lot of customers find PaaS complicated for use and instead find container based services Docker much easier to use. Darrow further adds that one primary reason why PaaS may appear to be DOA is because very few enterprises feel a real need to migrate to a wholesale PaaS solution since a lot of these businesses already own the infrastructure to interact with their application. Darrow’s article quotes James Watters from Pivotal challenge this. Watters feels the low demand for PaaS from companies is more of a supply constraint and less of demand. He expects PaaS to grow and become more established over the next two years.
All said and done, one thing is clear - companies today are still slow in adopting PaaS. According to Mike Kavis from Cloud Technology Partners, there are more than a couple of reasons why this is the case. He says that the first and foremost reason for the slow adoption rate is the confusing marketing message from PaaS vendors; including the likes of Microsoft and Google. Kavis emphasizes on the need to differentiate between the various PaaS offerings like public PaaS, hybrid PaaS solutions, domain specific solutions and PaaS-like IaaS features provided by AWS Elastic BeanStack. Beyond this, Kavis also believes that a lack of a mature product and the feature limitations from the current crop of PaaS solutions are other reasons why businesses still have not invested in a PaaS solution.
Given the present state of affairs, 2014 could be a make-or-break year for PaaS as a cloud service category. As James Watters predicts, we could see PaaS mature and grow over this year and next to establish itself as a category distinct from SaaS and IaaS. In case that does not happen, we could, in all likelihood, mourn the demise of PaaS over this year. Whatever the case, one thing is clear - PaaS may not exist as a solution category. But the features and solutions it proposes are something that the industry terribly needs and even if PaaS doesn’t exist, we will continue to see its usage as features provided in IaaS and SaaS solutions.