The current situation and future of PAAs analysis

Although the investment is huge and has been through so many years of development, but so far, PAAs still does not attract many customers. This article outlines the views of several analysts on the status of PAAs and its future.   Over the years, PAAs has been viewed as a necessary "glue" between the application/service/software layer and the cloud infrastructure layer, as Paul Miller in the article "PAAs is dying?" The view:  the platform is clear, persuasive, and powerful. It should be a fundamental part, much more interesting than a bunch of cheap virtual devices running on commercial hardware. It should be the driving force of the cloud and the reason that the cloud can continue to transform the enterprise and business models around the world.  miller added, but "PAAs as a category is far from meeting this expectation". Recently, someone has begun to wonder if we really need Paas,paas is it going to die or is it being swallowed by Laas? Indeed, when it comes to Gartner's recent application PAAs (Apaas) quadrant, Nancy gohring notes that the number of PAAs customers is staggering. Google has about 30,000 customers, mostly small sites, and several big names, such as Snapchat and Khan Academy. Although no specific figures are provided, gohring that Salesforce "may have more customers than Google App engine" because "Google is ranked higher than Salesforce", but the good news is only these:  What makes Gartner's report more interesting is that fewer of those customers are suppliers. That's because some of them are very famous. For example, Gartner writes, IBM's SCAS service is estimated to be less than 50 customers worldwide. Gartner points out that the number of SAP customers is less than 100. According to Gartner, Red hat, as the focus of the OpenShift service, "almost no paying customers."   New participants are clearer: "Cloudcontrol, a German company that serves the European market, says he has 400 paid clients." Docker has 500 paying customers. "  Seven years later (according to Google Trends, the PAAs term first appeared on the Web in March 2007), PAAs is not as mature as Laas (see Amazon AWS) or as SaaS (see Salesforce). 451 Study Group recently titled "PAAs is becoming a feature of Laas?" "The question is whether PAAs is being swallowed up by Laas," the report asked. In fact, at the beginning of becoming a PAAs provider, Google and Microsoft have already used several Laas features, CPUs, and storage extensionsIts products. How's  paas? Why is that? Can PAAs survive? Krishnan Subramanian believes that PAAs has just passed the hype peak, "reached a new level of maturity", and despite the deficiencies, but 2013, the PAAs domain has an important development: differentiation into two types:   PAAs by service orchestration--"Early vendors such as Google APP engine provide this type of PAAs and build platforms by combining the different services needed for application deployment. ”· PaaS by Container orchestration--"Docker is a typical example of a fast, lightweight Linux container that makes it easier for users to migrate applications between different cloud providers. Unlike virtual machines, which are only abstract from the original computer, containers can encapsulate the entire application and application environment. " mike Kavis lists three reasons why companies use PAAs slowly:  1, market turmoil-" there are public PAAs providers such as Heroku, Google, and Microsoft, as well as Apprenda, OpenShift, Pivotal, WSO2 and many other proprietary and hybrid PAAs solutions, as well as specific areas of PAAs solutions, are dedicated to disparate areas such as mobility, large data, social networking, and more. ... Laas providers also offer a number of services similar to PAAs, such as AWS Elastic Beanstack and other services. Even more puzzling, Google and Microsoft now offer LAAS services with their PAAs services. There's no one day I don't see people trying to compare Microsoft Azure and AWS, the former PAAs and the Laas. What's worse, even the ' experts ' can't differentiate correctly. " 2, lack of maturity-" the challenge of PAAs is that many customers already view SaaS and Laas as trustworthy services, and they gain more trust from within the enterprise (or at least for AWS). When AWS returned to PAAs in 2008, most of its clients were start-ups and small and medium-sized enterprises. There are also some impressive business success stories, but few. " 3, limitations--" There is a view that you simply pounce on these PAAs solutions and start coding. That may be true, but if you want the code to work, you have to understand the limitations of PAAs and build on it. For example, you can consider a Heroku dyno as a container that holds all the infrastructure and stacks. Heroku will randomly recycle dyno when it feels right., and it only gives you 10 seconds to process the error code ... So, all of a sudden, your code starts with a very specific target PAAs platform, creating a locking pattern that you probably don't want ... The amount of work required to address the limitations of most PAAs architectures, and the cost of consuming so many abstract resources, makes it very difficult to adopt and expensive and sometimes impossible to adopt. Because Laas does not limit resources like PAAs, very large applications can get better services from Laas solutions. "  in the article for Forbes, Ben Kepes writes that there are too many competitors in the PAAs market, not yet unified:  in short, this is a special period for PAAs. Pivotal one this type of cloud computing seems to occupy the vast majority of attention, giving other cloud providers a scratching of how to deliver differentiated services. At the same time, Redhat is trying to achieve some kind of rapid breakthrough for its PAAs version openshift. Similarly, stalwarts Heroku (now owned by Salesforce.com) and Engineyard have been adjusting their PAAs development direction. In this case, some OpenStack competitors have decided to create their own PAAs solution Solum, so the customer is facing a confusing and confusing market. Aside from the behemoths of Seattle, AWS and Microsoft, there is, on the surface, one supplier for every six companies in the world that have actually decided to buy PAAs.   When it comes to the future of PAAs, there are various insights.  brandon Butler in the article "is the PAAs market really dying as we know it?"  tim Crawford, a CIO strategy advisor at his independent company Avoa, said: "Many people will say that PAAs is dead." "He believes that even with greater Laas and SaaS providers taking action, PAAs as an independent hosting platform is still a difficult business to adopt."  kepes: "It's time for PAAs to really prove itself not just a very wonderful concept in the marketplace." " kavis:" PAAs does not die: In fact, it is only in the initial stage, is waiting for a storm in the market. The problem is that, in the current form of PAAs, many companies still don't want to put chips on top. " subramanian:" Premature announcement of PAAs death will ultimately hurt an industry, or it may derive enormous value from its continued development. "  After the hype, whether 2014 will become a PAAs mature reunification of the year-for enterprises, it is enough to represent aReliable solutions and, conversely, suppliers can take the opportunity to push the PAAs towards success--yet to be seen.