April 4, 2013
In a couple of weeks I’ll be heading to Portland, Oregon briefly to attend past of the OpenStack summit (disclosures – the OpenStack foundation is supporting my T&E to attend the summit and Diversity Limited is writing a general guide to open source clouds for the foundation) As part of the upcoming summit, today marks the release of the latest version of the project, codenamed Grizzly. Now it’s fair to say that the OpenStack initiative has come in for some criticism in the past – the governance process was a little shaky to start with, the project was initially dominated by one vendor, and early releases of the product where a little flaky. But the benefit of the open source model is that it encourages rapid iteration and improvement and Grizzly, the seventh release, shows the value of a community that includes some 500-plus contributors and, in the case of the Grizzly release, 7620 individual patches. It’s also a reflection that OpenStack is starting to se some pretty active adoption – some notable users include Best Buy, Bloomberg, Comcast, CERN, HP, NeCTAR, NSA, Rackspace and Samsung. As a foil to the “but it’s simply a Rackspace initiative”, the foundation reports that 45 individual companies contributed to this release.
So, what’s new. There are over 200 new features in this release across the different main parts of OpenStack. Some highlights from the release announcement include:
- OpenStack Compute – Compute delivers improved production operations at greater scale, with “Cells” to manage distributed clusters and the “NoDB” host architecture to reduce reliance on a central database. Improvements in virtualization management deliver new features and greater support for multiple hypervisors, including ESX, KVM, Xen and Hyper-V. Additional functionality was added for bare metal provisioning, shared storage protocols and online networking features such as the ability to hot add/remove network devices
- OpenStack Object Storage – Cloud operators can now take advantage of quotas to automatically control the growth of their object storage environments. Additionally, the ability to perform bulk operations makes it easier to deploy and manage large clusters and provides an improved experience for end users. Cross-origin resource sharing (CORS) enables browser connections directly to the back-end storage environment, improving the performance and scalability of web-integrated object storage clusters
- OpenStack Block Storage – The second full release of OpenStack Block Storage delivers a full storage service for managing heterogeneous storage environments from a centralized access point. A new intelligent scheduler allows cloud end users to allocate storage based on the workload, whether they are looking for performance, efficiency, or cost effectiveness. The community also added drivers for a diverse selection of backend storage devices, including Ceph/RBD, Coraid, EMC, Hewlett-Packard, Huawei, IBM, NetApp, Red Hat/Gluster, SolidFire and Zadara
- OpenStack Networking – The leading network-as-a-service platform enables advanced network automation, allowing users to control their networking technology of choice. Those choices grew tremendously with the Grizzly release, with the addition of support for Big Switch, Hyper-V, PlumGrid, Brocade and Midonet to complement the existing support for Open vSwitch, Cisco UCS/Nexus, Linux Bridge, Nicira, Ryu OpenFlow, and NEC OpenFlow. OpenStack Networking achieves greater scale and higher availability by distributing L3/L4 and dynamic host configuration protocol (DHCP) services across multiple servers. A new load-balancing-as-a-service (LBaaS) framework and API lays the groundwork for further innovation from the broad base of networking companies already integrating with OpenStack
- OpenStack Dashboard – OpenStack Dashboard brings an improved user experience, greater multilingual support, and exposes new features across OpenStack clouds, like Networking and LBaaS. The Grizzly Dashboard is also backwards compatible with the Folsom release, allowing users to take advantage of additional features in their Folsom cloud prior to a full upgrade to the latest version
- OpenStack Identity – A new token format based on standard PKI functionality provides major performance improvements and allows offline token authentication by clients without requiring additional Identity service calls. OpenStack Identity also delivers more organized management of multi-tenant environments with support for groups, impersonation, role-based access controls (RBAC), and greater capability to delegate administrative tasks
- OpenStack Image Service – There were major advancements in image sharing between cloud end users, and the creation of a set of common properties on images to provide more discoverable images and better performance when retrieving images
I’ll have far more to report in a few weeks after the OpenStack summit. For the moment I’ll take this release at face value – it hasn’t been without its trials and tribulations but it seems that the foundational concept of OpenStack, to clone the community, success and momentum of Linux to a cloud operating system, was a sound one. No one knows where this will all lead but it’s hard to see how OpenStack, with it’s massively broad industry support, won’t be part of the cloud landscape for a long time to come.