vege - Fotolia
Over the years, OpenStack has received a high degree of interest and attention but relatively low levels of adoption. It's notoriously difficult to deploy and many say it's still missing pieces, sowing doubt that it'll ever live up to expectations, six years after it was started. This month we're asking our Advisory Board experts whether they think OpenStack will live up to its initial hype, whether it will slowly fade away or if it will remain relevant only to a select few. In short, what's the future of OpenStack?
Brian Kirsch, Milwaukee Area Technical College
Brian KirschMilwaukee Area Technical College
A lot had been made of the open source cloud creation platform OpenStack. Depending on your news sources, OpenStack is set to take over the cloud. Several hosting companies, and even technology competitors, have some type of relationship with OpenStack and the platform has several public champions within the open source community. With a product seemingly so ready to take the cloud by storm, it really hasn't. In fact, overall it has a low rate of adoption. As we all know, the cloud isn't going away, even if private cloud deployments have slowed a bit. While powerful, OpenStack has a weak spot; it's difficult to get it up and running. How difficult, though, is dependent on your background and training, but it's sufficient to say that an OpenStack installation is more involved than a simple wizard.
While the online guide is comprehensive, it can also require very precise and large numbers of command-line steps to navigate, and input from the installation guide to get up and running. While the complexity of the installation doesn't affect the performance or features of OpenStack, it does have a large impact on its install base and adoption rate. In IT, we need to do more with less and if your installation takes a small army of people or an inordinate amount of time to complete, is savings of an open source platform really worth it? Everyone wants to save money with licensing, and hopping on the open source bandwagon seems like the right answer, but we have seen this before. Remember Linux on the desktop, and look where that ended up.
I am not against open source, and I have high hopes for the future of OpenStack. Unfortunately, those hopes are also grounded in the reality of many very smart people who have tried and failed to get OpenStack up and running. While the dedicated Linux/OpenStack fans will argue about the complexity, the simple truth is that the installation and configuration is not geared for most IT people. That doesn't mean you need to make your product fool-proof, but when you release a product, you have to make sure the audience you are targeting can work with it. If OpenStack is going to grow -- or even survive -- it has to cater to a much wider audience than just its hardcore fans. That doesn't mean investing in even more pages of documentation, it means taking a step back to really look at what it takes to get OpenStack up and running.
Jim O'Reilly, Volanto
OpenStack is a very attractive solution for end users seeking to build an inexpensive cloud and for system vendors transitioning to a services model. It has been, and remains, well supported in the industry, but it has grown dramatically in scope. The result is that the level of completeness varies among the 30 or more projects that make up the total package. This creates pitfalls for the unwary, since architecting a long-term cloud platform around a core with options that are still in early development is risky.
The OpenStack Foundation is tackling the issue of completeness with OpenStack Navigator. This online tool provides an assessment of completeness for each module plus links, detailed specs and descriptions, making life for users easier. It also provides use case models in a variety of configurations.
From a big picture perspective, OpenStack looks to be the best approach for private or hybrid clouds, despite -- or even perhaps because of -- its rapid evolution. It is addressing real issues of security, databases and big data in a structured way. The long-term vision looks about right.
There is however a dark cloud on the horizon for the future of OpenStack. The lack of wide area network bandwidth threatens to jeopardize the hybrid cloud, since the crucial transfer of data between the public and private cloud segments will be a bottleneck for the foreseeable future. Add to that the lack of easy installation tools, just now getting remedied, and it seems some IT shops are deciding either to go completely public. Another factor to consider is the efficiency boost that containers bring to the table, which might disrupt the economics of a public cloud connection and push the all-private cloud approach.
It isn't yet clear if the negative factors will derail private cloud efforts enough to change the market direction, or if the boost from containers will help the future of OpenStack. In my view, this won't happen. I think OpenStack will be perceived as a platform for private and hybrid clouds, simply because the industry is too heavily invested in it already.
An introduction to OpenStack components
Step-by-step OpenStack installation guide
Is an OpenStack IaaS the best choice?