New OpenStack Queens release provides support for GPUs, containers to meet edge, NFV and machine learning workload demands



The OpenStack community released on Wednesday Queens, the 17th version of the open source cloud infrastructure software. A packed release resulting from a six-month development cycle, Queens offers advancements benefiting for both enterprises with mission-critical workloads as well as organizations investing in emerging use cases like containers, NFV, edge computing and machine learning. The software now powers 60 public cloud data centers and thousands of private clouds at a scale of more than six million physical cores.

The Queens release also includes significant enhancements to support emerging use cases such as machine learning, NFV, container integration and edge computing. In Nova, vGPU (virtual graphic processing units) support lets cloud administrators define flavors to request specific resources and resolutions for vGPUs. End users can boot VMs that have vGPUs, an important capacity for graphics-intensive workloads and many scientific, artificial intelligence and machine learning workloads.

Cyborg, a framework for managing hardware and software acceleration resources such as GPU, FPGA, CryptoCards and DPDK/SPDK, debuts in the Queens release. Acceleration has become a necessity rather than an option, particularly for telcos with NFV workloads. With Cyborg, operators can list, identify and discover accelerators, attach and detach accelerators to an instance, and install and uninstall drives. It can be used standalone or in conjunction with Nova or Ironic.

It also delivers instance repair—long available for VMs in Nova—is now available for bare metal instances in Ironic. Operators can now troubleshoot misconfigured bare metal nodes or recover from issues like a lost SSH key, an important enhancement since production usage of Ironic has jumped from 9 to 20 percent between April and November of 2017

OpenStack is the preferred platform for containers deployed in private cloud, and the community continues to expand microservices features in Queens. Kuryr adds a CNI daemon to increase the scalability of operations on Kubernetes. To support HA, the CNI daemon watches for pod events, eliminating the need to wait on the Kubernetes API for each event. Pods can be created even if the controller goes down.

Two new projects, OpenStack-Helm and LOCI, described in more detail below, support edge computing applications such as those presented in the white paper “Cloud Edge Computing: Beyond the Data Center” recently published by members of the OSF Edge Computing Group.

The three most relevant things for edge are OpenStack-Helm, LOCI and acceleration support, which will be a requirement in a lot of edge environments. Helm containerizes OpenStack services and puts them in a Kubernetes pod. (You can see how this works with a live upgrade demoed by AT&T at the Sydney Summit).

LOCI is a set of lightweight container images, the project takes a thinner approach to the container – offering a smaller, much more portable, lower footprint. Then operators can add configuration to the outside and use Kubernetes to manage the overall service functionality.

Together they make it easier to deploy a lot of OpenStack environments and to upgrade in an automated way as well as operate in a zero-touch way over the long term and with a smaller footprint than a traditional datacenter OpenStack deployment. Those attributes will all be very important in edge deployments.

Zun is a new OpenStack project that allows users to start and run containers without having to manage servers or clusters. It seamlessly adds advanced enterprise networking, storage and authentication capabilities to containers by integrating with Neutron, Cinder, Keystone and other core OpenStack services.

Enterprise adoption of cloud continues to expand, and by next year, 60 percent of enterprise workloads will run in the cloud, according to 451 Research’s Voice of the Enterprise: Cloud Transformation, Workloads and Key Projects survey.

451 Research also reports that enterprise adoption of OpenStack is expanding in parallel, with enterprises in nearly all verticals and regions now running mission-critical workloads on OpenStack software. To support these workloads, the Queens release includes robust, enterprise-oriented features, most notably the multi-attach feature in Cinder.

Cinder Multi-Attach enables operators to attach the same Cinder volume to multiple VMs. If one node goes down, the other takes over and has access to the volume. This redundancy—which supports high availability (HA) for mission critical workloads—is one of the most-requested features in cloud environments and has remained a largely unmet challenge in computing until now.

“In the early days of cloud, the use cases were fairly narrow, but the scope of cloud use cases today has expanded to include a massive variety of workloads,” said OpenStack Foundation Chief Operating Officer Mark Collier. “Just as the scope of cloud evolves, OpenStack evolves as a platform. Consider the vGPU and containers enhancements in the Queens release: they address opposite ends of the use-case spectrum and demonstrate how OpenStack has the flexibility and ability to support changing open infrastructure requirements.”

Leave a Reply

WWPI – Covering the best in IT since 1980