PaaS, containers, and serverless are creating a multi-platform world
It is the era of multi-platform cloud computing. But what is “multi-platform”? In this article, Chip Childers, CTO of Cloud Foundry Foundation, gives us a tour of what’s included in this new era and explains why most companies will be, or are already, adopting a multi-platform strategy.
In 2016, we spent a lot of time talking about “multi-cloud” and the need for technologies that enable businesses of all kinds, but particularly enterprises, to use different cloud providers as part of their business strategy. As this reality has taken hold, we’ve seen a similar wave of adopting multiple abstraction models for application platforms. With both horizontal (multi-cloud) and vertical (multi-abstraction) optionality, we have entered the “multi-platform” era of cloud computing.
So, what is “multi-platform”? It’s a logical extension of the multi-cloud concept. In essence, it means organizations are taking full advantage of the cloud-native technologies at hand. The Cloud Foundry Foundation, since its inception in January 2015, has been a strong supporter of an ecosystem of great technologies, and this new era is the latest development to benefit enterprise IT departments and organizations.
I’ve been in the enterprise technology business for two decades now, so I’m really not surprised by all this optionality. People in enterprise technology must meet specific demands when it comes to business, and a bespoke solution comprised of multiple technologies tweaked to accommodate their needs is a no-brainer. It’s like ordering a pizza for your family – one of your kids only eats cheese, one likes pepperoni, and you like mushrooms and olives, so you order a pie with three separate toppings, divided in thirds, to be enjoyed in tandem. That’s what we found in our latest research as well – folks are ordering a “pizza” to fit the needs of their organization.
We published our findings in a report entitled “Where PaaS, Containers and Serverless Stand in a Multi-Platform World.” It can be downloaded from our website. The top level finding is that IT decision makers across the world are using PaaS, containers and serverless in tandem:
- 64% of respondents are now using a combination of PaaS and containers
- 43% are using a combination of PaaS and serverless
- 42% percent are using a combination of containers and serverless
- 39% are using all three technologies
It’s clear that enterprises are moving dynamically toward PaaS, containers and serverless, and I think it’s safe to extrapolate this growth means a majority of companies will be, or are already, adopting a multi-platform strategy.
The case for Platform-as-a-Service, or PaaS, now goes back several years with the advent of cloud computing as a serious option within enterprise IT. The original conceptualization of cloud computing as Infrastructure (IaaS), Platform (PaaS), and Software (SaaS) as services has been blurred over time, but retains some validity. PaaS in particular has emerged as the foundational technology upon which the modern enterprise is embracing software development.
Adopting a PaaS platform, such as the Cloud Foundry Application Runtime, gives the developer an application-centric abstraction to work with, allowing an enterprise to focus on the business requirements without significant changes to their application’s architectural approach (or at least less changes than required for the “serverless” model). While PaaS platforms should be designed for ease of operations, the key to their value lies in the focus on the developer experience.
As the most widely-recognized PaaS, the Cloud Foundry Application Runtime is in use by more than half of the Fortune 500.
SEE ALSO: How do we keep containers secure?
Containers, or more specifically Linux containers, have taken off in the enterprise as well. Largely, the industry has consolidated around the Kubernetes open source project as the preeminent tool (and therefore API) for deployment and management of container fleets. Nearly all multi-abstraction platforms offer a container-as-a-service experience, and most provide that through Kubernetes.
Containers have a dual benefit. First, for developers, they offer a larger degree of “portability” to move software between environments with a high degree of confidence that it will function in each new environment. Second, for operators and CIOs, containers have the potential for significant cost savings through the use of less system resources than traditional virtualization provides (this is because the containers share a kernel, while virtualized systems share hardware). That said, nearly every container deployment runs on top of fleets of VMs. These two attributes have made container platforms the new defacto standard for ISV software delivery.
The Cloud Foundry Application Runtime (CFAR) itself has container technology built into its very DNA from the beginning, as one of the earliest platforms to adopt the Linux kernel features that make containers work. More recently, the Cloud Foundry community has been building this second platform abstraction through the Cloud Foundry Container Runtime. This is a distribution of the Kubernetes container orchestration tool that is deployed and managed by the same operational tool chain used for CFAR.
Enterprises are drawn to containers for their ability to move entire runtime environments among operating systems and cloud infrastructures. As the name implies, containers let you seal off your precious cargo and transport it without worrying about losing it or breaking anything along the way.
Cloud computing, as outlined by Nicholas Carr in his influential 2008 book, The Big Switch, has the potential to be measured in real-time units, as we do with water or electricity. We’re getting close to that goal of “true utility” and serverless is the most relevant current example of this.
Rather than signing the monthly and annual contracts as found with yesterday’s legacy technology, the public-cloud pioneers were able to offer instances of computing resources on an hourly basis. Now, with serverless, customers can be billed by the second or even by each execution of a function. Users need not know or care what actual servers are is providing this resource, so the service looks “serverless” from the customer’s perspective.
Serverless is especially suited for event-driven applications and services, such as those driven by Internet of Things (IoT) devices that may be sending very small amounts of data over brief time periods. The other key area where serverless is valuable is serving as a kind of “scripting language” for our new cloud-scale operating environments. It is still early days for this abstraction, but we are seeing it grow in popularity and use at an extraordinary rate.
All together now
Our report found that IT decision makers rank “integration with existing tools” and “flexibility to work with new tools” as Top Five attributes, along with security, support, and price. The three abstractions discussed here bring those attributes to the discussion. Add to that the practicality of platform, effectiveness of containers, and efficiency of serverless, and we can see why these three technologies are increasingly being used in various combinations.
The digitization of industry, supply and value chains, communications networks, and business and government services will no doubt continue to pick up pace, in my view. I look forward to seeing how great teams and companies find their own unique paths to success.
Our new report tells me this is going to happen. The report has too many other great data points and results for me to provide them all here, so I would urge you to download it and find out for yourself what our respondents told us. You will see why these three technologies serve as the foundations for this new, multi-platform world.
This article is part of the latest JAX Magazine issue. You can download it now for free.
Have you adopted serverless and loved it, or do you prefer containers? Are you still unsure and want to know more before making a decision? This JAX Magazine issue will give you everything you need to know about containers and serverless computing but it won’t decide for you.