Open source PaaS ready for prime time?
Perhaps the most satisfying thing at events such as the recent Cloud Foundry Summit in Basel is seeing things you’ve been writing about in abstract terms, and predicting for pretty much the last five to 10 years, actually happening for real.
Because the fact is, open-source cloud computing works. Not only can you build a local Platform-as-a-Service (PaaS) that runs elastic applications, spinning up new VMs or containers in the background as needed, but you can run that PaaS on your own Infrastructure-as-a-Service (IaaS), like OpenStack. If need be you can also push those elastic apps to PaaS on a public cloud IaaS such as Microsoft Azure, Google Compute Platform (GCP) or Amazon AWS. If you are building cloud-native apps, this is the state of the art.
Yes, there are bumps along the road: speakers at CF Summit spoke of issues with scaling, or with legacy monitoring tools, or simply – and most often – with persuading sceptical colleagues that it really will work. But they also talked about how they solved those problems and how they fed the fixes back into the project to benefit others – this being the norm for open-source, of course. And they shared ideas on how to deal with and win over the sceptics.
It was similar a couple of weeks ago when sitting down with Red Hat, whose OpenShift platform is one of the other main open-source options for PaaS. Interestingly, while Cloud Foundry and OpenShift might look like direct competitors at first sight, the fact that some user organisations run both suggests that while they do cover much of the same ground, their strengths lie in different areas.
To put it simply, Cloud Foundry is probably the better choice if you are developing new born-on-the-web apps, while OpenShift has strengths in other areas of the enterprise, for example where you need to connect legacy apps to the web or cloud. OpenShift also appears to be slightly ahead on its hybrid cloud capabilities, whether that’s apps ‘bursting’ from private to public, or building apps that combine private and public resources or services.
So while you could do it all with a single PaaS, many organisations choose to use more than one – and of course these aren’t the only open-source options around. As well as several Cloud Foundry siblings or derivatives such as IBM Bluemix, Fujitsu K5 and SAP Cloud Platform, there are PaaS frameworks based on Apache Mesos and GCP. (On the closed-source side, the likes of AWS, Azure and VMware vSphere all have PaaS capabilities too.)
Don’t get hung up on the terms, though. PaaS, IaaS and even CaaS (containers-as-a-service) are all just terms for automating various levels of an application stack – the important thing is they all concern automation, because automation is what web apps and the cloud are built upon.
If some or all of that application stack is open-source, so much the better. That’s not so much because it’s free, as in zero purchase cost – it can be, but organisations are increasingly opting for commercially-supported versions. Rather, it’s because the code is open, whether to performance tuning, bug-finding, integration with other tools, or whatever, and it’s because it is shared. That all helps to build supportive communities among its users and it adds flexibility, and that’s good news for modern software developers.
Bryan is a technology enthusiast and industry veteran. He has been analysing, explaining and writing about IT and business in a highly engaging manner for around three decades. His experience spans the early days of minicomputers and PC technology, through the emergence of cellular data and smart mobile devices, to the latest developments of the software-defined age in which we all live today. Over his career, Bryan has seen at first-hand how IT changes the world – and how the world changes IT – and he brings that extensive insight to his role as an industry analyst.