Saturday, October 29, 2005

Hell is frozen but SUN is shining

IBM has agreed to sell Sun's operating system with its BladeCenter servers in "the coming months," according to an IBM spokesman. This is quite the surprise given IBM's contentious relationship with Sun. IBM's services organization, however, does do a large amount of business selling Sun servers and Solaris, which may have helped seal the deal. Sun killed its line of blade servers after a horrible go at the market, which IBM and HP lead.

Sun has long promised that a major OEM would back its Solaris x86 push. But up until now, the major vendors largely mocked Sun's Unix embrace. Jonathan Schwartz, president at Sun, was happy gloat about the move.

"Im pleased to announce we've signed up our first tier 1 systems vendor as a Solaris supporter: it's IBM, and their decision to provide comprehensive support for Solaris on Bladecenter definitely puts them ahead of the other blade vendors in offering a truly OS neutral product," he said in a globule. "As a result of our agreement, IBM will be adding value to BladeCenter, optimizing Solaris for IBM hardware offerings, adding volume to the Solaris community, and proving that the best choice for customers is, in fact, real choice.

Thursday, October 27, 2005

Firms gather to promote grids in Boston

IT giants including IBM and Sun were out in force in October promoting the business benefits of grid computing, announcing new partners and customers, and launching tools to make it easier to deploy the technology.

At the GridWorld conference in Boston, IBM inked partnerships with Absoft, which has designed new development tools to aid grid deployment; and with business intelligence specialist SAS Institute, which launched an automated grid management capability.

Tuesday, October 25, 2005

Who is using Oracle Grid?, Everybody says Larry Ellison!

LARRY ELLISON, chief of enterprise database giant Oracle, stood in front of thousands of customers at Oracle OpenWorld in San Francisco last month and told them they would eventually all migrate to enterprise computing grids.

He repeated himself in the style of a preacher: "If you were to ask me what percentage of our users will be using grids, I'd say everybody."

Then he paused to search eyes in the audience, daring a response.

It's not clear what he saw, but he didn't try to make the point again.

Oracle has been very vocal in promoting the technology, which promises high-computing power on low-cost non-proprietary hardware.

While it has gained acceptance in academic and scientific research environments, however, mainstream enterprises are reluctant to put it to the test.

IDC storage research director Graham Penn estimates only a handful of enterprises in Australia have installed grid computing technology.

"Some of the universities, of course, are playing with it, and occasionally you get it in a financial modelling environment but you're not typically seeing it all in a commercial environment."

Grid computing has been given a number of technical and marketing aliases, such as utility computing, clustered computing and on-demand computing.

These have generated some confusion among potential users.

Most industry experts are content to strip away the marketing-speak and adopt a simple definition: lots of computers working together to act as one.

Scientific and engineering professionals began using grid computing in the early 1990s.

The ground work for wider adoption in commercial computing didn't begin until late in that decade.

Under pressure to support the emerging e-commerce sector, in which business outcomes had been unpredictable, the computing industry sought ways to match the sums its customers were investing in technology more closely with their levels of business activity.

Oracle first released its grid technology, called RAC (Real Application Cluster), in 2000 with the Oracle 9i database.

It partnered with computer maker Dell to develop grids made up of low-cost Intel-Linux operating platforms.

Sun Microsystems and IBM, which previously relied on proprietary Unix-based hardware for revenue, came up with their own versions of grid computing.

IBM developed its On Demand strategy of running applications across heterogenous hardware environments on its xSeries of servers, which include the Intel-IBM Power 5 platforms.

Sun developed grid software that allowed AMD Opteron-based systems to be pooled.

It released a pay-per-use grid, so organisations could buy processing time as they would electricity or gas.

The latest version of Oracle's database, 10g, has been certified to multiple databases and applications for up to 100 processors or nodes.

IBM computing business development executive Andrew Brockford says grid computing developed a lot of industry hype.

It may have been "oversold", he says.

"Like a lot of emerging technologies, there's a lot of hoopla and there's a lot of discussion that has been divided.

"Perhaps the concept of the grid has been somewhat oversold by both business and academia," Brockford says.

Sun Microsystems server specialist Robert Becker says the technology can't replace traditional single-servers in every instance. "The reality is that, yes that works for some customers, but a lot of customers went down that path, even here in Australia, and it just did not pan out the way Oracle said it would," Becker says.

Two of Sun's big Australian customers spent a lot of money on failed grid computing projects, he says.

Oracle, which has Qantas as a grid database customer, says Sun's comments are FUD, industry parlance for fear uncertainty and doubt.

Oracle Asia-Pacific vice-president Roland Slee says grid technology isn't "good news" for players that stand to lose in a move from proprietary single-server systems.

"It is very controversial technology because, traditionally, managing large databases required that you have a large computer," Slee says.

"Computers have a high cost and they are also high-margin.

"By virtue of their affordability, grids are very attractive to customers, so they're not always good news for everybody in the marketplace."

The take-up of Oracle's grids in Australia has been impressive with more than 60 customers using RAC deployments of low-cost Linux-Intel server platforms.

Oracle's partner, Dell, is less bullish when it comes to discussing the technology.

"It's fair to say we've seen large-scale grid demonstration projects take place in universities and research institutions, but the take-up of grid technologies in enterprise data centres was always going to follow a different pattern," Dell Asia-Pacific enterprise marketing and alliances manager Justin Boyd says.

Penn says it is a case of "horses for courses".

Grid computing has had some success in financial modelling, but it is more likely to be used in research and academic environments that have lots of spare hardware at hand to experiment with, he says.

It's a courageous CIO who's prepared to try to retrofit the technology to the company's critical business processes, Penn says.

"You've got to be reasonably brave if you're going to be a pioneer; if you're going to change the way you're doing things," he says.

Grid projects work best when the applications run on them can easily be broken into small chunks, he says. "In some cases that's easy because there are millions of very small chunks and it's not a problem," he says.

"If a lot of data has to flow through the networks, maybe you've got to have big pipes."

Brockford says the uptake of grid computing is higher where applications lend themselves to the technology.

However, Sun says it's a straight case of horses for courses. "For certain types of workloads and for larger organisations you need larger vertical servers with more CPUs because there's just no way they can scale to those computing outcomes," Becker says.

Slee says he hasn't come across any situation where Oracle's grid computing can't be applied.

Oracle recently added the NSW Office of State Revenue to its customer list. Slee says the OSR replaced a $1.5 million Unix-based set-up with a $150,000 grid system and doubled its processing performance. Oracle says deployments of its RAC in pure research settings are in the minority, and most of its education customers use it for commercial applications.

Most deployments of Oracle's grid technology locally are in the government sector, followed by finance and telecommunications.

Martin Power, chief information officer of Holmesglen TAFE college in Victoria, has been managing an Oracle grid since 2002.

One of the key advantages of grid system architecture is the ability to provide resources to applications as they are required, he says.

"Where you have a single server and it hits its peak, you just wait until it finishes," he says. "In a grid model you can add extra capacity to that particular function for the time it requires."

Brockford says there is potential for financial institutions to use grids' provisioning ability to cope with peaks and troughs in demand on systems.

For portions of this story concerning Larry Ellison, Andew Colley flew to Oracle OpenWorld, San Francisco, as a guest of Oracle.

Tuesday, October 18, 2005

LinuxHPC and Cluster Builder offer more Services, Again

Ask the Cluster/Grid Expert;

“This site is an important key in discovering information about cluster and grid computing,” said Michael Jackson, president of Cluster Resources, Inc. “It educates people on the different options available and directs them to the solutions that work best for their circumstances. Adding new content on current HPC trends is critical for the site to continue being a valuable research tool.”

“Ask the Cluster/Grid Expert” enhances the Cluster Builder research by starting a valuable knowledge base of questions and answers relating to clusters and grids. Anyone can submit a question about clustering or grid computing and the Cluster Expert will seek out the answer through means such as industry experts, mailing lists and other research. Also, Cluster Builder visitors can share their knowledge and expertise with others in the community by responding to the questions.

Other popular services of Cluster Builder consist of the community submissions tool and the Request For Quotes (RFQ) service (http://www.linuxhpc.org/quote). Community submissions allow visitors to suggest sites that expand Cluster Builder’s research base. The RFQ service allows those purchasing a cluster to submit one request-for-quote form that is then sent to the vendors of their choice, instead of having to discover and go through the process with each vendor individually.

“The features on Cluster Builder, like LinuxHPC.org’s RFQ service and the new Ask the Cluster/Grid Expert, help expand its research capacity giving people greater quality information to help them make better decisions,” said Ken Farmer, editor-in-chief of LinuxHPC.org
More info @ PRLEAP

Friday, October 14, 2005

New GridSolve Middleware Release

A powerful new version of GridSolve middleware has just been released by the Innovative Computing Laboratory (ICL) of the University of Tennessee, led by Jack Dongarra. GridSolve is a more highly evolved form of NetSolve, the popular Grid middleware system widely known for its deployability, its ease of use via familiar interfaces -- such as Matlab, C and Fortran -- and its integration with popular Grid technologies. GridSolve represents a new implementation of NetSolve's core functionality that is based on the emerging standard for Grid Remote Procedure Call (GridRPC). It not only makes Grid computing even easier to deploy and use, it incorporates major enhancements based on real world experiences and feedback from the NetSolve user community.

Sunday, October 09, 2005

GlobusWORLD and the Global Grid Forum Joins GridWorld

FRAMINGHAM, MA, October 2005 – IDG World Expo, the leading producer of world-class tradeshows, conferences and events for technology markets, today announced that GlobusWORLD™ will become a part of GridWorld™, scheduled to take place September 10-14, 2006 in Washington, D.C. GridWorld will now incorporate complementary content from two leading organizations in the grid computing market: GlobusWORLD, focused on the Globus Toolkit as the leading open source Grid infrastructure software, and the Global Grid Forum (GGF), a community of users, developers and vendors leading the global standardization effort for grid computing.

GRIDWORLD is the premier B2B conference focused on the commercial benefits of grid computing for Enterprise IT strategists, worldwide.

Complete Press release is here

Friday, October 07, 2005

Microsoft's Hey say Hey ! Grid here we come

Microsoft is Not going to be left behind like it did with the internet few years(decades) ago. It is working on the grid initiative. But it is also going where the money is, rather than computing grids (which is the flavor of the time), it is focusing on Data Grids.

Microsoft is creating a "Cluster Compute" version of Windows and intends to work more closely with grid industry standards bodies, Tony Hey, the company's corporate vice president of technical computing, said in an interview with CNET News.com on Tuesday.

Wednesday, October 05, 2005

News release: Moab Access Portal® Source Code Available

To improve Moab's ease-of-use and expand administrator control, Cluster Resources released Moab Access Portal source code to its customers and partners. This new benefit lets organizations configure the Access Portal to accommodate unique site-specific needs. For example, organizations using Moab can now create automated batch submission on the Access Portal so end users don't have to learn batch codes, which decreases training requirements for administrators.

With available source code, organizations can also integrate Moab Access Portal with their existing Web portals. End users can use the Web portal they are already familiar with and organizations can consolidate programs to simplify management.

Moab Access Portal source code is a value add of Moab Cluster Suite®, Moab Grid Suite® or a Cluster Resources' support contract. Please contact Cluster Resources at (801) 873-3400 to learn more about downloading Access Portal source code.

Monday, October 03, 2005

Cisco announces new datacenter architecture for grid computing

With an eye toward providing an open utility architecture, Cisco Systems (Profile, Products, Articles) last week unveiled its new compute networking and virtualization architecture, VFrame. The architecture combines a set of InfiniBand-based SFSes (server fabric switches) with the VFrame virtualization software suite. Although much of this release finds its roots in Cisco’s acquisition of TopSpin, the company has gone to significant lengths to rebrand and extend the technology in time for this announcement.
Even so, VFrame constitutes a significant step forward in terms of utility application deployment. By offering the VFrame library of APIs, Cisco makes it easier for developers of existing apps to port their software into an effective utility/grid model.

And by using VFrame as the “glue” between utility deployments and their hardware layers, Cisco offers one of the first intervendor utility and virtualization management suites in the industry. It is not strictly heterogeneous, as managed devices still need to be VFrame-compliant, but within its boundaries, it has one of the widest vendor libraries, including Altiris, EMC, Hewlett-Packard, Microsoft , Oracle , SAP , Tivoli.

Sunday, October 02, 2005

Clusters, Grids And Virtual Organizations

In effect, grids are distributed collections of servers that can be assigned to work on different parts of the same task. Unlike the servers in clusters, which are typically dedicated to a particular application or set of applications, the servers in a grid may be allocated to perform different applications, or parts of applications, at different times of the day. Not surprisingly, there are many more clusters in use today than there are grids.

In grid terminology, the group of processors, memory, and storage that forms a grid is also called a virtual organization (VO). Each VO has policies describing how resources are discovered, scheduled, secured and paid for. Grid resource managers implement and enforce these policies. Applications and resources that have the software and interfaces to participate in the VO are called grid-aware or grid-enabled.

Advertisement 3
The operational trade-off with a grid is taking on another layer of management in order to make better use of your computing Latest News about computing resources. There is no rule of thumb for grid cost savings yet, but the Enterprise Grid Alliance (EGA) and other grid proponents are working on ROI studies to demonstrate that grids can increase server utilization and reduce the total number of servers.

Different vendors and consortia have different names for grid resource managers, but their functionality is similar: The servers and other resources tell the resource manager how much processing, memory or storage they have available and the workload characteristics that they can handle. Then the resource manager allocates resources to applications that request them. The grid standards-setting body, the Global Grid Forum (GGF), has done much research in this area, and a set ofopen source Latest News about open source tools is available from the grid community Globus Alliance.

Truly the technology is tricky, but a dedicated I.T. organization that sees the grid deployment as strategic should succeed. As with all I.T. initiatives, the keys are simple, but not easy -- a clear strategic vision, policy and process.

Big vendors like EMC Latest News about EMC, HP, IBM Latest News about IBM, Microsoft Latest News about Microsoft, Oracle Latest News about Oracle and Sun are in the forefront of promoting grid technology, but early adopters also are using software from companies such as United Devices and Platform for simpler grid applications: to steal unused processing power from desktop PCs to process batch jobs.

Think about how much time you spend in meetings or on the phone, when your PC could be doing other jobs. The simple grid concept is appealing, but there are some drawbacks, not the least of which are security Latest News about Security and stability concerns. And, as always, standards and interoperability represent a double-edged sword: Industry leaders must work for standards, but they also want to maintain proprietary product differentiators.