Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage News OpenCompute and OpenStack Span Hardware and Software Cloud Infrastructure and Platform Solutions

OpenCompute and OpenStack Span Hardware and Software Cloud Infrastructure and Platform Solutions

This item in japanese

Last week Facebook launched OpenCompute in collaboration with OpenStack with the goal of building energy efficient computing infrastructures at a low cost. Yesterday OpenStack announced the Cactus release which includes API extensions, improved features and support for enterprise technologies to enable larger scale cloud deployments. In a separate announcement, GigaSpaces in collaboration with Citrix released their PaaS/SaaS enablement platform to OpenStack.

In the same spirit as open source software, OpenCompute shares the data center and server specifications and drawings that Facebook developed in a year. These specifications are the basis of Facebook's Prineville, Oregon data center that is 38% more efficient while costing 24% less in comparison with their other facilities. Here are some energy efficiency features of the Oregon data center:

  • Facebook’s energy consumption per unit of computing power has declined by 38%.
  • The new data center has a PUE of 1.07, well below the EPA-defined state-of-the-art industry average of 1.5. This means 93% of the energy from the grid makes it into every Open Compute server.
  • We've removed centralized chillers, eliminated traditional inline UPS systems and removed a 480V to 208V transformation.
  • Ethernet-powered LED lighting and passive cooling infrastructure reduce energy spent on running the facility.

While Jonathan Heiliger, VP of technical operations at Facebook said:

The ultimate goal of the Open Compute Project, however, is to spark a collaborative dialogue. We’re already talking with our peers about how we can work together on Open Compute Project technology. We want to recruit others to be part of this collaboration -- and we invite you to join us in this mission to collectively develop the most efficient computing infrastructure possible.

Marco Arment, founder of Instapaper interpreted it as an attempt to attract engineering talent. At the launch, OpenStack announced its collaboration with Facebook to ensure that their stack can run on top of OpenCompute's reference architectures. 

OpenStack is making progress with their third major release: Cactus, 10 weeks after the Bexar release. Since Bexar, 4700 commits and 40 new features were added by developers at Canonical, Citrix, FathomDB, Grid Dynamics, NTT Group and Rackspace across three projects: OpenStack Compute(Nova), OpenStack Impage Registry and Delivery(Glance) and OpenStack Object Storage(Swift). Here is a set of features from a press release on talkincloud:

Taken directly from the official press release, here’s the new features that Cactus brings to each, starting with OpenStack Compute:

  • Support for all major virtualization technologies, including Microsoft Hyper-V, KVM, LXC, QEMU, UML, VMware vSphere, Xen and Citrix XenServer;
  • Live Migration support for KVM, which enables users to move running virtual machines from one physical host to another;
  • Enhanced network management features, including IPv6 support for flat networking, advanced network configuration and QoS management with XenServer;
  • New OpenStack API 1.1 with support for extensions, which allow developers to innovate more quickly by adding extensions to their local OpenStack installations ahead of the code being accepted by the OpenStack community as a whole;
  • Multi-cluster region support, which allows administrators to manage servers in clusters, and create fault zones and availability zones; and
  • Support for enterprise storage solutions, including Solaris iSCSI and HP SAN.

New features in OpenStack Object Storage include:

  • An enhanced authentication system;
  • The ability to collect and serve data that enables integration of service provider billing solutions or internal chargebacks; and
  • Static web serving for faster and easier access to content.

And the OpenStack Image Service gets a command line interface and image verification.

While OpenStack is focusing their efforts on the Infrastructure layer, GigaSpaces wants to:

extend the openness and flexibility of OpenStack to the application platform layer, enabling enterprise customers to choose a cost-effective cloud solution that will not only allow them to automatically scale to meet the demands of employees and customers, but also evolve with their business needs and ensure that they are not locked in to a long-term solution.

The solution, which is an integration between Citrix's Open Cloud Solution(Citrix® NetScaler®, Citrix® XenServer®) and GigaSpaces eXtreme Application Platform(XAP), will provide the following benefits:

Openness:  An enterprise-grade hybrid cloud stack solution that includes OpenStack-based IaaS and PaaS and is open to a wide variety of applications, including those based on Spring, Java, .Net, and other non-Java platforms.
Flexibility:  Develop applications with a choice of application stacks (Tomcat, MySQL, NoSQL, Hadoop, Ruby, etc.) and handle provisioning, elasticity, multi-tenancy, continuous high availability, and monitoring of those services in a consistent way across the stack.
Integration: By leveraging the Citrix OpenCloud solution GigaSpaces will provide tighter integration between the platform and the infrastructure layers to up-level services offered through OpenStack clouds.
Scalability: Enterprisescan use GigaSpaces as the enterprise PaaS layer specifically geared for big-data analytics, e-commerce, and financial applications.
SaaS-Enablement: Ability for SaaS ISVs to SaaS-enable applications while delivering the same solution off- and on-premises.

The ability to switch to any cloud provider is achieved through an integration with the JClouds project that provides a common abstraction platform as depicted in the logical architectural diagram shared by Nati Shalom:


Nati in his blog post also talks about the enhanced platform driven monitoring features in this offering that handles failures and scaling automatically and provides a holistic view of the application and infrastructure to the Ops team.



Rate this Article