Working with OpenStack storage: Tips on Cinder, Swift and the cloud

Last updated:April 2014

Editor's note

OpenStack, a series of open source storage, cloud and computing projects, had a hard time getting a solid footing in the world of enterprise infrastructures largely because of its do-it-yourself nature. Many IT professionals were and still are wary about implementing OpenStack because it takes a special skill set to work with the technology, and should anything go wrong, there's no vendor support available. However, its DIY aspect is what lured IT administrators to try OpenStack. Open source platforms are customizable; if anything doesn't work well in a specific environment, an administrator can change the code to fix the problem. In addition, because OpenStack storage isn't tethered to a vendor, IT departments can save money by using commodity hardware.

As OpenStack continues to develop, more enterprises are using it as a platform to build their private clouds, including well-known companies such as Bloomberg and Vendors are increasingly getting behind OpenStack projects as well. Inktank, for example, sells a commercially supported version of Ceph, a unified systems option for OpenStack storage. Red Hat, along with Dell and Hewlett-Packard (HP) to name a few, also sell their own distributions of OpenStack storage.

Even so, working with OpenStack storage and OpenStack-based clouds can be overwhelming for any IT professional. This guide will help with the decision to transition to an open source platform. You'll learn the basics of OpenStack storage, tips for working with OpenStack block and object storage (code-named Cinder and Swift, respectively) and how the technology fits into an open source cloud implementation.

1Classifying OpenStack storage: Cinder and Swift

Currently, OpenStack storage comprises block and object storage options while an OpenStack file system project is in the works. There are, however, still specifications and compatibility concerns to be aware of. OpenStack Block Storage, also known as Cinder, can only be used in environments employing OpenStack compute. OpenStack object storage, or Swift, not only has its own application programming interface (API) but supports the Amazon Simple Storage Service API. The following links dig a little deeper into how Cinder and Swift work and explain how they're best implemented.

2Working with an OpenStack-based cloud

It's becoming more common to see OpenStack storage in private cloud implementations, and OpenStack APIs are now competing with market leaders for use in public cloud environments. But, as with any open source technology, many IT shops lack the skills needed to effectively deploy and work with it. To solve that problem, some vendors have begun leveraging OpenStack technology to create Private Cloud as a Service offerings.

3Latest updates from vendors leveraging OpenStack

Part of OpenStack's growing popularity is due to increased vendor support. Among the top players is Red Hat, which sells its own distribution of OpenStack that includes management software and a partner program; it also launched an educational initiative with Intel aimed to accelerate OpenStack adoption. In addition, HP, IBM and Inktank offer OpenStack-based products of their own. Find the latest developments below.