OpenStack's new release, Havana, is the eighth major release of this open cloud-infrastructure project. Havana emerged from beta last week with tasty new features.
You can think of OpenStack as a soup where the ingredients come from many major enterprise software players. Hopefully the taste will improve enough to attract enterprises and even public cloud providers that have hesitated to take a nibble.
[ Get the no-nonsense explanations and advice you need to take real advantage of cloud computing in InfoWorld editors' 21-page Cloud Computing Deep Dive PDF special report. | Stay up on the cloud with InfoWorld's Cloud Computing Report newsletter. ]
InfoWorld's Serdar Yegulalp has provided a great overview of what's new in the Havana release, including:
- The previous release of OpenStack, Grizzly, had load-balancing functionality, but Havana expands on it and adds firewall as a service.
- OpenStack sports a component named Heat, a new addition to Havana, that can talk to other components through their APIs and use templates to organize how they're put together and how they behave.
- Havana features a metering and data-collection framework (code-named "Ceilometer"), originally devised for customer billing but since expanded to include functions like alerting, for a broad range of data-collection drivers for big data systems (such as HBase and MongoDB).
As I've pointed out before, the hype around and vendor interest in OpenStack has not driven the desired level of enterprise adoption. In large enterprises I deal with, people like the concept of OpenStack, and some are using the distributions. However, I see a few bumps in the road in terms of implementations and even some missing features. In general, enterprise users want to wait for both the standard and the distribution instances to bake a bit longer.
Although the new features, such as APIs for orchestration and use-based accounting, are needed, I have a fundamental problem with Havana: There should be more in such a major release. That said, a standard IaaS platform with so many cooks in the kitchen will always be slow to evolve. I'll withhold my final judgment until this release is in production.
Whatever is finally delivered, we'll see vendors that sell OpenStack distributions add the missing features -- so that they can sell more OpenStack distributions. That means more nonstandard features, moving further away from OpenStack's core concept of being a standard. So far, that fracturing beyond the core has not been an issue, so maybe it won't become one in Havana.
Overall, OpenStack is a good thing. The market likes to use technology based on standards, and OpenStack seems to fit that bill. But at the end of the day, it has to deliver features that enterprises expect. It's time for OpenStack to hit the accelerator.
This article, "It's time for OpenStack to hit the accelerator," originally appeared at InfoWorld.com. Read more of David Linthicum's Cloud Computing blog and track the latest developments in cloud computing at InfoWorld.com. For the latest business technology news, follow InfoWorld.com on Twitter.