Bladewatch.com

http://www.theengineer.co.uk/sectors/electronics/news/grid-computing-helps-visualise-motion-of-seismic-events/1008044.article

Researchers in Taiwan are planning to use grid computing to visualise the motion of earthquakes after they occur. They hope this will cut the time of creating ’shake movies’ from a few hours to just minutes, providing valuable information to rescuers once an earthquake has occurred.

As animations that show the ground motion of seismic events, shake movies provide information as to where the strongest shaking has occurred, helping to ensure rescue efforts and resources are directed to where they are most needed.

An interesting article illustrating the possibilities of hpc or grid computing, it’s always great to see how technology is enriching lives or creating new opportunities, do check it out.

http://www.prnewswire.com/news-releases/platform-computing-announces-support-for-mapreduce-118834534.html

SAN JOSE, Calif., March 29, 2011 /PRNewswire/ — Platform Computing, the leader in cluster, grid and cloud management software, today announced the company is now bringing enterprise-class distributed computing to business analytics applications that process “big” data using the Apache Hadoop MapReduce programming model. Based on more than 18 years of industry leadership in workload management for high performance computing (HPC) applications, Platform Computing’s analytics solutions are a natural expansion of the company’s distributed computing experience built on the company’s core technologies, Platform LSF and Platform Symphony.

It’s always great to see and hear about innovation of grid computing and it’s movement into the cloud, which by its very nature can make grid computing more accessible to different user groups, simply buying in the capacity you need, or even selling the capacity you have spare can create real opportunities for new services and options for revenue generation. I’m off to read up more.

From one of the CIOs that I regularly speak with:

“About five years ago it was multi-core or performance that they spoke about when trying to sell you a server, then we moved to the green message, the performance per watt message, and now it’s cloud, multi nic, multi core and large memory support…”

A rather skeptical viewpoint, but one that remains valid and which raises some issues around not only how we approach the cloud and virtualization message but how we continue the dialogue with our end user community both in the business and technology communities. The focus should remain on end user empowerment, providing the right tools and technologies in order to deliver business benefits, providing insight and examples, concepts that could be used in order to deliver or form parts of a solution.

At the same time we need to see cloud as the opportunity that it is, absolve ourselves of comments around ‘it’s outsourcing online’ or ‘buying in service’ to thinking of it as a vehicle to not only provide infrastructure or services, but as a platform that can create opportunities for revenue generation, opening new markets and users to be the platform that joins the dots, just as the internet has proved to be through email, through community, rich media like youtube, would I have discovered Giles Peterson on Radio 1 if I hadn’t listened to him on the iPlayer if it meant staying up to 3am when his show is on?

Cloud is the next generation, the next concept and re-invention of IT services, products and solutions as IT re-invents itself every decade or less what represents the most significance to me though is the accessibility and the opportunities that we see presenting themselves. Virtualization brought us the possibility of having more servers but using less hardware, grid brought us real opportunities to achieve high performance and low latency compute farms using shared infrastructure and absolutely delivered more with less, the job that might take overnight on one pc might take minutes or hours on the grid. Cloud though is more exciting because it opens opportunities for the Small business, for the one man band, the manufacturer and the enterprise at different levels, and presents for the vendors ways of supplying products or services in a format and at a price that might not have previously been possible – online backup, email as a service, salesforce.com or googlemail, virtual infrastructure from Amazon for example. It’s these things that excite me, that the garage with twelve employees that might have previously had to pay thousands of pounds for the setup and on going maintenance of their IT might instead buy it as a service and level the playing fields between organizations, the social media, the email, the web services, backup and tools that the national dealer might be available to the stand alone dealer, creating both opportunities for the small businesses in agility and business to business services, whilst giving the enterprise the chance to realign their business and IT needs, focusing their time on making their inhouse systems work and buying in the services they need without the associated barriers to entry that exist in the traditional outsourced model.

The startup can now subscribe to an email service from google or using Exchange, have their own web servers virtually from Amazon, have their compute or dev farms and online storage and backup all provided down a wire reducing the start up costs, the barriers to entry real and perceived. We therefore need to be discussing cloud not only as a concept of how you as a vendor are supplying devices cloud ready, or how I as a customer need to get cloud ready, but how cloud works, how it can create opportunities for all, how it works for your business and how to leverage it and take it from an abstract concept, one which means many things to many people, to a concept that works for you and your business.

Last of all it’s important for vendors and service providers (as well as customers) to continue dipping their toes in both worlds, the virtual down-the-wire cloud space, and the legacy world. There are many companies both service providers, enterprises and consumers adopting cloud loving the agility and on demand presence that it gives them, but equally there remain enough of us legacy people with simple and complex problems which need addressed. We need to be adjusting the message, the support and the empowerment we provide to the target market, for that small business talking about their first server (which might just be an up scaled desktop) talking cloud is great, but they aren’t there yet, they might be given the right conditions, given the right set of need for services and solutions, but in the meantime is it a 1u a 2u or a Microserver you need, and do you need a guy to plug it in and set it up. Let us not forget it’s the small start up operating out the garage that could be your next cloud provider, your next Facebook or google, help them with their problems and their opportunities and the rest will follow, removing the barriers to entry, making it just that little bit easier can be the difference between a one of server sale and a partnership leading to years of commitment both in products, services and the cloud solutions you are working on for now and the future…

http://www.newswise.com/articles/grid-and-cloud-computing-meet-at-globusworld-2011

Newswise — (Chicago, IL) February 22, 2011 – Researchers, software developers, and cyberinfrastructure providers will convene at GlobusWORLD 2011 to see and experience new technologies at the intersection of grid and cloud computing. GlobusWORLD 2011, to be held April 11-13 at Argonne, near Chicago, will feature Globus user success stories, expanded software-as-a-service capabilities in Globus Online, updates on Globus Toolkit development, and sneak peeks into new features in the pipeline. For more information and to register, visit the GlobusWORLD web site (www.globusworld.org).

This does sound like an interesting event particularly as it discusses both the grid and cloud computing space, both platforms which continue merge as we start offering a range of services around the HPC and compute on demand space, as well as virtualization or applications down a wire so to speak. There are exciting times ahead and real opportunities for innovation, for transformation and revenue generation, not to mention making applications and services more accessible where they might not necessarily have been so before thanks to the economics involved, as we scale up the users the opportunities to reduce licensing costs, to pay on use create models of consumption at a lower entry cost, Office on demand so to speak, rather than the full edition so to speak, great for the user and the enterprise alike.

http://www.platform.com/news-coverage/2011/jpmorgan-wins-afta-award-for-best-green-it-initiative

Datacenters represent 20 percent of JPMorgan’s total energy bill and 80 percent of its carbon footprint, so it was a natural place to begin green investments. By implementing a global internal cloud for risk management, JPMorgan achieved a 50 percent reduction in the compute capacity required for derivatives risk computation. The 14-month project, which cost less than $1 million, allowed the bank to save $6 million in 2009 and reduce its carbon footprint by 4,500 tons of carbon dioxide per year.

One of my colleagues had sent me this reminding me about the conversation that we’d had many years ago, about harvesting compute resources both in a follow the sun capacity and by using the available compute capacity from nodes that were idle. It’s great to see JP Morgan achieving this, and fantastic to see some operational and financial benefits of doing so, think of the possibilities in an enterprise space, where we could harvest capacity from the non production or production systems that have spare capacity or aren’t being used outside of business hours, very very cool, do check out the pdf.

I got this emailed to me last week and I just replied, I thought in the interests of openess I’d include my reply for all to see and correct me if necesary.

Hi Martin,

The IT guys are telling me we should look at the HP SL6500, what is it and how is it different from the DL range of servers?

Thanks

Bill

IT Architect

Companyx

My reply is below:

Hi Bill,

Thanks for your mail, it’s always great to get emails. Anyway, the SL is HP’s scalable server platform which is designed for high scale out environments. That is it’s a server that is designed to be deployed in volume, providing the maximum number of nodes ina given footprint but not quite going down the blade route which some customers have felt was not quite right for them. HP sometimes refer to it as their ExSO or Extreme Scale Out server.

They have two ranges the SL6000 and SL6500, both with different selling points and features, but with common features, in that you buy a tray upon which you can have two or more servers in that tray dependent on the configurations that you require, the number of disks, the u-size etc.

Are they a suitable alternative to DL servers? Briefly, yes but it’s a platform that is designed for volume deployments, and might not necessarily deliver the flexibility that you have with your existing rack servers.

IBM offer a similar product which is equally good and worth looking at, they refer to it as the IBM System x iDataPlex, both platforms have their benefits and are better in some areas than others, so which you choose to examine is going to depend on your business, your IT and where you are in the server space.

There’s also Dell’s cloud server solution, their C series rack servers which were designed with cloud in mind, they’re interesting rack servers and feature some unique and very cool features.

Hope that helps, if you need anything, do get back to me.

Regards

Martin MacLeod

Martin237@gmail.com

Bladewatch.com

http://www.bladewatch.com/2010/01/08/follow-the-moon/

I was reading an interesting article talking about virtualization of storage being the next big thing, and for many it will be. I wonder though if we should not be looking more towards the follow the moon vehicle for IT infrastructure, the concept of data center virtualization.

We need to be thinking of virtualization as an evolving path, in which we continue to further abstract the end user from the infrastructure and the application, where we move towards service down a wire or online rather than locking the user to a specific device with a client application, with all the anciliary components to deliver that service. At the same time from an IT cost and business empowerment angle virtualization of the data center allows us to transform not only how we organize and support the IT services that the business needs, it allows us to look at how we host and power these services towards a follow the moon approach.

I was having a chat with colleagues over dinner the other night about follow the moon/follow the sun, they had some great comments and got me thinking, I’ve written about it before but thought I’d publish this brief PowerPoint document that I put together sometime ago. It’s in PDF here.

It’s meant as an overview and I confess that it could be missing some concepts, but nonetheless do check it out.

Hope it helps, if you have any questions, do email me.

Regards

Martin

http://www.platform.com/press-releases/2010/platform-computing-makes-hpc-in-the-cloud-a-reality

• Through its cloud computing solutions, Platform Computing helps improve utilization and resource capacity while helping organizations avoid the costs and delays of large capital purchases.  The company offers three separate paths for HPC users to dynamically scale resources to handle peak workload necessary for HPC applications:

• Solution 1: Integrated Cluster with the Cloud (Platform LSF) – Uses Platform LSF’s dynamic host capabilities along with a cloud service, such as the Amazon Virtual Private Cloud (VPC), to provide cloud resources that appear to operate within the HPC datacenter (local IP addresses, host names, etc.).  Platform LSF recognizes outside cloud services as valid resources and integrates the resources into the HPC infrastructure.  Workloads can then be sent to the cloud portion of the HPC infrastructure to run applications as designated by policy or on a job-by-job basis.
• Solution 2: Multi-Cluster to the Cloud (Platform LSF with Platform MultiCluster) – Allows users to start up a new cluster in any cloud or hosting provider environment without a dedicated link such as Amazon VPC.  Used in conjunction with Platform LSF, Platform MultiCluster enables users to distribute workloads between internal HPC clusters and external cloud resources, according to policy.
• Solution 3: Dynamic Cluster Extension to the Cloud (Platform LSF with Platform ISF) –  Combined with Platform LSF, Platform ISF creates a powerful and flexible cloud management solution that dynamically scales clusters to include resources within the data center infrastructure or externally to a cloud provider based on the application workload.  Platform ISF includes resource kits to manage physical or virtual servers, multiple operating systems, virtual machines and cloud resources.

It will be interesting to see what applications can benefit from these concepts and services, anything Platform and other vendors can do to meet business requirements, whether through enhancing the ability to scale up or right size the infrastructure on-demand has to be a good thing. I’m off to check out more.

http://www.platform.com/press-releases/2010/platform-computing-debuts-platform-lsf-8

NEW ORLEANS, SUPERCOMPUTING ‘10, November 15, 2010 – Platform Computing , the leader in cluster, grid and cloud management software, today announced the availability of the latest version of its flagship product family, Platform LSF, the industry’s most comprehensive workload scheduling solution for high performance computing (HPC).  Designed to handle complex, distributed HPC environments, Platform LSF maximizes existing IT infrastructures by allowing more work to be done with fewer computing resources in the fastest time possible.  Version 8 boasts a number of new features including the ability to delegate administrative rights to line of business managers; live, dynamic cluster reconfiguration; guaranteed resources to ensure service level agreements (SLAs) are met; flexible fairshare scheduling policies; and unparalleled scalability to support the large clusters in use today.

It’s great to see the innovations and improvements in Version 8, the ability to delegate administrative or management rights to the end user community, to the business or application manager is not only a welcome functional improvement, it can actually help with the on-boarding and integration process. That as a grid manager I can grant the relevant rights, give elements of control, of management and monitoring to the end user is an exciting concept, and helps further the message of grid being an open platform for delivery, for business empowerment. Anything that can be done to aid in low latency configurations, in achieving or exceeding service level agreements has to be a good thing, to provide grid/hpc as a platform for delivering application compute resources and end user functionality, very cool, I’m off to read up more.

http://www.bladenetwork.net/?pageid=1684

SANTA CLARA, CA, October 14, 2010 – BLADE Network Technologies (BLADE), the data center switching company, announced today the RackSwitch G8264™, the first single-chip 40 Gigabit Ethernet (GbE) top-of-rack switch to deliver more than one terabit of low-latency throughput to the data center.

BLADE’s delivery of the RackSwitch G8264 marks the first time a single-chip switch is available for terabit-scale deployment of 10GbE. Aimed at fueling mainstream adoption of 10 and 40 Gigabit Ethernet for the enterprise data center, the RackSwitch G8264 is priced at an industry-pacesetting $350 per 10GbE port with 40GbE connectivity at just $1,400 per port.

With 64-10GbE ports, up to four-40GbE ports and 1.28 terabits of non-blocking throughput, the RackSwitch G8264 is designed for HPC clusters, cloud computing, algorithmic trading and other I/O-intensive and highly virtualized workloads. Its ultra-efficient single-chip switch fabric ensures deterministic “fair” latency and consistent throughput across all port combinations. This new top-of-rack switch extends BLADE’s award-winning RackSwitch family that brings unprecedented speed and intelligence to the edge of the network.

RackSwitch G8264 advantages include:

• 1^st single-chip Ethernet top-of-rack switch exceeding 1Tb bandwidth
• 1^st top of rack switch with true 40GbE
• Best-in-class data center features, including VMready for VM-aware network virtualization
• 1^st standards-compliant, non-proprietary TRILL platform

I wonder if this kind of solution might bring further opportunities in the hpc and electronic markets space where ultra low latency seems to be the aim of the game, even small differences in results or transactions can be the difference between making money and not, acquiring that deal and not. Anything the vendors can do to address this using a combination of the right range of technologies both software and hardware has to be a good thing for the end user community and in illustrating IT as being a business empowerment tool.

Lossless, deterministic low-latency, low power, low cost