Ten years ago, he said, most researchers worked with a high-end PC and hoped there was enough disk space to run their applications.Now, Howard says, you run your compute and have your data stored in a national facility, you get more compute than you have in the local environment, and then you pull your results back across the network.As someone whose history stretches back to Australia's early Internet, with involvement in early OpenFlow developments, and three-and-a-half years after the NCI first investigated OpenStack, Howard has an insider's view of OpenStack's effectiveness today.That's a big thing for the NCI, because its scientific HPC is characterised by very big but relatively short-lived data flows, a large number of users, and facilities all over the country.The drive to commodity into networking equipment, vendors offering low-cost switches, all combined with an SDN control plane, offer significant savings, he said, but it's the sophistication that's demonstrated the importance of SDN to the NCI.Rather than being a painfully difficult environment to use, Howard said, the NCI's attention with OpenStack has been trying to balance which features we make available at what time, how we take advantage of those features, and how to train the researchers to take the best advantage of that environment.
The federation of service providers and software vendors will begin stocking its virtual shelves from June 7Xavier Poisson, Hewlett Packard Enterprise's vice president for hybrid IT in Europe, Middle East and Africa, at a meeting for Cloud28 partners near Amsterdam on May 19, 2016Cloud28 , the cloud services federation backed by Hewlett Packard Enterprise, now wants to help you install enterprise applications, not just choose them from its catalog.The federation plans to open its new App Center for business later this summer, and will begin stocking its virtual shelves on June 7 with the opening of an App Onboarding Center.Once strictly an open source, OpenStack shop, it is now embracing Microsoft Azure, VMware, Ormuco and Docker.One of the advantages of Cloud28 for participating service providers operating in only one country is that it makes them more visible to independent software vendors from other countries looking for new distribution channels, said Khaled Chaar, managing director of German cloud service provider Pironet NDH.Although many applications are already cloud-ready, for a typical business the 20 percent or so of the applications it uses that aren't cloud-ready are probably the most valuable, industry-specific ones, he said.Making it easier to move those to the cloud will present significant advantages, he said.
EnterpriseDB's latest play to become a leading enterprise database vendor has seen the company reveal the EDB Postgres platform.The open source vendor has chosen to embrace a wide ecosystem of vendors and technologies in order to show that open source is the way forward for enterprise databases.Linster said: "Instead of saying all data has to live in the same infrastructure, we understand that sometimes data is better off sitting elsewhere."In addition to the software connections that EDB is making, it has also developed a partner ecosystem with leading hardware vendors such as IBM, HPE, and Dell.New capabilities that have been added include cloud management support for private cloud use, particularly OpenStack.Linster said that the idea is to simplify management and deployment and to get better access to other vendors, "we don't live in isolation," he said.
Cray has always been associated with speed and power and its latest computing beast called the Cray Urika-GX system has been designed specifically for big data workloads.It also includes its own graph database engine called the Cray Graph Engine, which the company claims is ten to 100 times faster than current graph solutions running complex analytics operations.While the customer still deals with applications built on top of the platform, Cray will handle all of the big-picture stuff and work with the customer s IT department on the rest.While it s all well and good to say you ll take care of the software maintenance, it gets tricky when the customer is building stuff on top of the software you installed and the vendor is responsible to make sure it all works.Cray s Ryan Waite, senior vice president of products, insists that Cray has a long history of working closely with its customers and can handle whatever grey areas may arise.In other words, they have to compete, so the multi-million dollar price tags of yesteryear are long gone.
Given the rise of big data analytics as a necessity for businesses, the demand for tools and systems that can process data has risen, which is why Cray has revealed a new system.The Cray Urika-GX is being called the "first agile analytics platform that fuses supercomputing technologies with an open, enterprise-ready software framework for big data analytics," according to the company.The openness of the system continues with enterprise tools such as OpenStack for management and Apache Mesos for dynamic configuration."We took the Aries system interconnect from our supercomputers, the industry-standard architecture of our clusters, the scalable graph engine from the Urika-GD appliance, and the pre-integrated, open infrastructure of our Urika-XA system and combined them into one agile analytics platform."According to the company this technology is typically ten to 100 times faster than current graph solutions for complex analytics operations.The system features Intel Xeon Broadwell cores, 22 terabytes of memory, 35 terabytes of local SSD storage capacity, and the Aries supercomputing interconnect.
Supercomputer shop Cray is branching into analytics with tailored high-end systems packing open-source number-crunching software.Cray Tuesday unveiled the Urika-GX ticking some of the best-known technology boxes and due in the second-half of 2016.The Urika-GX is preinstalled with OpenStack and Apache Mesos for cloud and data-center abstraction, Spark and Hadoop for big analytics.Cray has also built its own graph engine software that ll run in a version of the system called Urika-GD Graph Discovery appliance.It claims this is 10 to 100 times faster than current graph processing systems and can run on anything from a single to thousands of processors without a hit on performance.There s 35TB of PCIe SSD on-node storage and Aries high-speed interconnect from Intel that debuted with Cray s XC30 super computer in 2013.
The architecture of the software is modular and offers numerous components, these include the likes of Nova for providing virtual machines upon demand; Swift is a scalable storage system that supports object storage; Sahara is a data processing service for OpenStack managed resources; and Neutron provides network connectivity-as-a-service between interface devices.OpenStack started as a joint project between Rackspace and NASA and is now managed by the OpenStack Foundation, a non-profit entity that was established in September 2012.Rackspace donated the code that powers its storage and content delivery service and production servers, while NASA contributed the technology that powers Nebula, which is an open source cloud computing program that provides on-demand computing power.Since its creation OpenStack has come in for criticism for its complexity, particularly related to installation.Adding to this problem has been a lack of skills available when it comes to installing and managing the system.The OpenStack community has placed a large focus on overcoming these issues with the creation of training and certification programs and work to simplify any complexity.
Tintri and its VM-aware storageThe company started in 2008 with the idea of providing a storage system particularly optimized for virtualized environments.At the beginning it was VMware-only, but now it also supports Microsoft Hyper-V and KVM RedHat and OpenStack with a similar set of functionalities.This is more like a federation of different storage arrays managed as a single system, and with the addition of smart data and workload placement features to help keep it balanced and efficient – resembling VMware DRS on the compute side.I m not saying it s perfect, though.They offer a seamless user experience and similar granularity to what Tintri promises.The VM-aware storage sales dilemmaVM-aware storage also has another problem: the buyer.
Even though NetApp has been showing less than brilliant financial results for a while now but who has among primary vendors?Excluding Ceph and without counting how many times Ceph is deployed for non-production workloads only , Solidfire is very relevant when it comes to OpenStack storage.Policies, APIs, QoS and many other details can help you do it all from vCenter, scripts or whatever software can use SolidFire APIs.The paradigm has changed drastically.Thanks to the introduction of Flash, IOPS and latency are less of an issue.They are working to make storage transparent and easy to consume from the upper layers of the stack – and without human intervention.
At this point in the project s development, the software isn t production ready and still being heavily developed.Its key tenet is to follow the self-proclaimed Google Infrastructure for Everyone Else or GIFEE approach to application development.Solutions were bespoke and managed in a structured manner to deliver reliability and high availability.The Architect s ViewTorus, to me, looks like another not invented here project that wants to re-invent something that s already out there and working.A better approach would have been to develop storage standards as part of the Open Container Initiative and allow existing storage providers to adapt or develop their solutions to an open and agreed standard.Over time Torus will probably be mostly deployed with CoreOS but I can t see it getting widespread traction, not with so many other competing solutions out there.
Network virtualization specialist Midokura today announced it has raised a $20 million Series B round with participation from Japanese fintech company Simplex and existing investors like Allen Miner and the Innovation Network Corporation of Japan.As enterprises move away from expensive proprietary networking hardware in favor of network virtualization and software-defined networking, Midokura offers a number of services that allow them to make this switch.The company s efforts mostly focus on the open source OpenStack platform which you can think of as an open source version of AWS that enterprises can run in their own datacenters .As Midokura CEO and co-founder Dan Dumitriu noted when I talked to him earlier this week, the company had barely gone to market when it raised its last funding round in 2014.As for the new funding round, Dumitriu told me that the current funding environment is definitely making it harder to raise money, so this round came mostly from insiders.This explains why the investors mostly hail from Japan, too, but Dumitriu noted that his company s strategic partnership with Fujitsu will hopefully help it capture more of the Asian market.
PodcastIt's El Reg's weekly tech-cast, hosted by Ed Saipetch and Ben Kepes.This week the dynamic duo discuss MPLS, open software, AWS and Bill Gates.Their special guest this week is Stephen Foskett, Organizer in Chief at Tech Field Day.The details… 1:00 HPE s Vegas Vacation 3:50 OpenStack Israel 8:06 Cisco MPLS fallout 11:19 Open software, commodity hardware and burning platforms 13:07 AWS outage hits down under and pizza riots 16:50 Bill Gates and a hundred thousand chickens 20:53 TBL: Decentralize the internet 26:22 FLASHBACK: The Clapper, 1984 26:50 Introducing special guest Stephen Foskett, Organizer in Chief at Tech Field Day 29:32 Media, analyst, blogger, evangelist, influencer or douchebag?31:43 Chuck is not John 38:45 Apple s WWDC and innovation 43:50 People want easy, not openListen with the Reg player below, or download here.Podcast Subscriber LinksFeed URL for other podcast tools †Juice, Zune, et cetera: http://nekkidtech.libsyn.com/rss
The Apache Software Foundation has announced that Libcloud 1.0, the cloud service interoperability Python library, is now generally available.Notably, version 1.0 introduces two new driver types, container-as-a-service and backup-as-a-service, with a full changelog available here – though the degree to which it will ease the burden on programmers is still not completely clear.Libcloud was created to provide a unified API so developers could better grapple with the swelling number of cloud services companies are utilising.Apache claims the library is currently used by more than 50 providers, including Amazon Web Services, Apache CloudStack, Google Cloud Platform, Microsoft Azure, OpenStack and VMware."When we first formed Apache Libcloud, there were many failed attempts to create a standard API specification for Cloud providers," said Tomaž Muraus, veep of Apache Libcloud.Originally developed and open-sourced back in 2009, by Muraus and colleagues at Cloudkick, the project joined Apache Incubator from which it graduated as a top-level project in 2011, days before the 0.5.0 version release.
OPNFV Project is driving uptake of NFV, but a lack of skills and manager commitment causes concernThe Open Platform for Network Function Virtualisation OPNFV Project is helping to drive the uptake of network-functions virtualisation NFV within the telecom industry, a new survey has found.HPE for example is backing the concept, which is similar in nature to software-defined networking SDN .The survey about the OPNFV Project was designed to gauge market perceptions of telecommunications network operators.Other key findings include the fact that 99 percent of respondents believe that OPNFV is relevant to upstream open source projects, with 85 percent of respondents ranking OpenStack as the most important upstream project to OPNFV s success, followed by OpenvSwitch 49 percent , KVM 42 percent and OpenDaylight 37 percent .Skills ShortageBut there remains some challenges, not least of which is a skills shortage, and lack of upper management support.Do you know all about the Internet of Things?
Hyper-converged - IT for complexity refugeesAiming to stem the Nutanix snd SimpliVity hyper-converged tide, HDS announced the UCP HC V240, its first hyper-converged UCP product, along with the entry-level converged UCP 2000, for its mid-market and enterprise customers.Think of UCP Unified Compute Platform systems as being like EMC Vblocks H-blocks?There is support for VMware, Microsoft and OpenStack virtualisation environments.HDS says it has tight integration with VMware operations and storage policy-based management SPBM , plus HDS data protection products.There are virtual machine VM centric pools of capacity, with consumption based on VM-level policies that can be changed on demand.Small and medium businesses are not in this listPore over this charted table to position HDS' converged and hyper-converged products in your mental landscape;We can surely expect a broadening UCP HC range from HDS after this toe dip in the water.
Juniper Networks seems to have big plans for its OpenContrail SDN controller: it would like to see it act as a kind of meta-controller for multiple cloud and data centre controllers.It's quietly pushed up the first code of what is, for now, dubbed project Ukai at Github.The alpha-level code has two top-line aims: to run global network resource management that gets translated to local resources in data centres; and to manage global identifiers like route targets and security group IDs.An example is a content distribution network: the operator needs to have the same virtual network configuration, network policy, and firewall rules everywhere.The effort aims to create a global controller that can support 3,000 read requests a second, 1,000 requests per second for configuration update requests, and handle up to 50 data centres.The current version of Ukai runs on OpenStack releases later than Liberty; Ubuntu 14.04 or higher, and OpenContrail 3.0 or higher.
News: Addition of support for Google Cloud Platform and further Azure integration aims to give customers more flexibility in company s hybrid cloud drive.The latest version of the platform, CloudForms 4.1, sees it now capable of supporting the three largest public cloud offerings in the market, Amazon Web Services, Microsoft Azure, and Google Cloud Platform.Expanding on the support for these clouds, the company has added support for software defined networking to bring network management for OpenStack networking in Neutron.The company has also added in greater automation capabilities with the integration of Ansible Tower.This should make it easier for users to provision complex applications and services through custom automation code.To further develop its portfolio of services, Red Hat has updated Insights, its SaaS analytics platform.
List: Containers, analytics, and a wedding, who said conferences were boring?On the analytics front the company advanced its portfolio of services with updates to Insights, its SaaS analytics platform.Insights will now offer full container workload analysis, updates OpenStack cloud capabilities, and an Actions Insight Planner to add customisation to task assignments.Red Hat has placed itself firmly in the market of hybrid cloud, and to achieve this it needs to be able to connect to other clouds.Basically this is a flexibility play saying to customers that if they want to adopt a hybrid cloud approach then Red Hat a wide reaching range of connections to the largest public clouds in the market.Drawing on historical examples of how inventors gave back to the broader community by not patenting their inventions, Whitehurst said: "Driving innovation means giving up control," said Whitehurst, followed by "innovation rarely comes from the boardroom or executive suite, it comes much closer to the customer."
Ubuntu funnels Hosted BootStack through e-shelter data centres for end to end private cloud managementCanonical and NTT Communications-owned data centre business e-shelter have launched a joint managed OpenStack private cloud service for e-shelter customers.The platform uses Ubuntu s Hosted BootStack , and is built on Ubuntu s own OpenStack distribution.Describing the product as end to end, e-shelter and Canonical will take care of the design, implementation and complete management of the OpenStack private cloud, which will run over dedicated hardware in e-shelter s European data centres.Hosted BootStack also takes advantage of Canonical s application modelling service Juju, alongside the company s Bare Metal as a Service MaaS – touting simpler provisioning, orchestration and management of cloud resources.Hosted BootStack clouds can be built on-premise or in dedicated colocation environments in e-shelter data centres.
The creation of data is exploding at a rate that threatens to bury businesses that don t have an adequate data strategy toward masses of structured and unstructured data.Mike Tuchen, CEO, Talend, told CBR that increasingly businesses are building ecosystems around a Salesforce, SAP, NetSuite, or Marketo.This may not be a particularly new development when it comes to the CRM and ERP markets but it is also now something that is happening on the analytics side.Tuchen said: Nowadays that market is shifting rapidly, more going to Hadoop and Spark and increasingly we are seeing interest in Amazon Web Services with their Redshift offerings and with Google with their Big Query and Google DataLab offerings, those are really the centre of mass systems are changing rapidly on the analytical side.Hybrid cloud is another trend that has been given a lot of air time, but for Tuchen this is to some extent still in its proof of concept phase.More broadly what I see customers want to do with an hybrid approach most of that today is about existing systems on premise connecting to things in the cloud, this concept of having true private clouds in the OpenStack like sense with AWS in public cloud is much a proof of concept level thing right now without really yet that broad mainstream adoption.
More

Top