Here's a great idea to put to your CIO: Why not run the company using a server operating system made by Mattel? It's the company behind Barbie and Hot Wheels (not to mention Tumblin' Monkeys), so it certainly knows a thing or two about toys. Maybe its designers have enough time to put together an enterprise OS.
Yeah, right. The idea is plain ridiculous, but is it any more ridiculous than using Apple's OS X Server or letting end users work on Macs in the enterprise?
Because the truth is, Apple is not really a computer company. It makes toys. It used to be a computer company called Apple Computer, but it dropped the "Computer" bit from its name in January 2007 as a tacit admission that it was now a consumer gadget maker, not to mention an online music retailer. Following the introduction of the iPhone and iPod Touch, two very pretty "boy's toys," the company's latest caper is the launch of its App Store.
The top-selling applications as I write are Band, Crash Bandicoot and Super Monkey Ball, which sounds uncomfortably similar — in name at least — to the aforementioned and very wonderful Tumblin' Monkeys.
Perhaps I am being unfair. After all, Microsoft makes toys too, and plenty of enterprises run their businesses using its server OSes. Just under 40 percent of all server spending was on Windows-based servers in the first quarter of 2008, according to IDC's Worldwide Quarterly Server Tracker. Yet Microsoft also makes the Zune, for example, as well as the legendary Microsoft Barney.
So why shouldn't enterprises take Apple seriously? Here's the problem: It can't walk and chew gum at the same time.
Microsoft is huge, and it is quite capable of doing more than one thing at a time. During the past two years, it worked on Vista, Windows Server 2008, the Hyper-V virtualization system and the Zune — all at the very same time.
The same cannot be said for Apple. It can certainly make great toys like the iPhone and iPod Touch, and there's no doubt it can create OSes. But, as it revealed last year, it can't do both at the same time:
"... iPhone contains the most sophisticated software ever shipped on a mobile device, and finishing it on time has not come without a price — we had to borrow some key software engineering and QA resources from our Mac OS X team," Apple announced. "As a result we will not be able to release Leopard at our Worldwide Developers Conference in early June as planned."
That's scary stuff and not what you want to hear from an OS developer. The iPhone software development effort wasn't a one off, either: The company has clearly been putting plenty of resources into developing version 2.0 of its iPhone (and iPod Touch) software during the past few months. Firmware for its consumer toys, not software for its computers, is the priority at the moment.
It's ironic, then, that many commentators have been suggesting the new iPhone 2.0 software (which adds features like support for Microsoft Exchange ActiveSync to make the phone more attractive to businesses) will raise Apple's profile in the business market and thus lead to a gradual increase in the use of Macs in the enterprise. Also ironic is a new industry group consisting of Atempo, Centrify, Group Logic, LANrev and Parallels to form the Enterprise Desktop Alliance (EDA). This consortium of software developers is dedicated, according to the group's Web site, to "making it easy to deploy, integrate and manage Macs in a Windows environment."
Running Macs in the enterprise doesn't seem like a good idea if Apple hasn't got enough engineers to provide the kind of resources an enterprise desktop OS inevitably needs. For example, if enough corporate bigwigs insist on bringing their shiny new MacBook Airs onto the LAN, you can be sure malware writers will start to target them. How fast will Apple be able to respond with patches if it's too busy selling Super Monkey Ball?
Granted, Microsoft isn't always the fastest to respond to newly discovered exploits, but there's no doubt it has the manpower to put to the task when it really wants to. Likewise, there's certainly no shortage or Linux and Unix patch-writing experts willing to devote their time to producing security fixes for their OSes.
So where does that leave Apple's server OS? In the nine years since its launch, it's gone precisely nowhere, and with Unix server spending declining in the first part of the year, according to IDC, it doesn't exactly look like sales are going to explode any time soon. Where will Apple be devoting its attention over the next year or two then: Developing its server OS or making more toys for the boys (and girls)? That's a tough one. Not.
Lest you think I am just another mindless Apple basher, I'll proudly admit to having an iPod Touch, a black Classic, white fourth- and third-gen models, a silver Nano, and three Shuffles (one lost) in a variety of colors. But would I put an Apple Server in my business? Not a chance.
In addition to writing for ServerWatch, where this column first appeared, Paul Rubens is an IT consultant and journalist based in Marlow on Thames, England. He has been programming, tinkering and generally sitting in front of computer screens since his first encounter with a DEC PDP-11 in 1979.
Saturday, August 9, 2008
Microsoft Serves Up SQL 2008
Microsoft today announced it's releasing to manufacturing SQL Server 2008, its enterprise database and business intelligence platform.
Originally planned for earlier this year to coincide with Windows Server 2008, SQL Server was delayed to insure the code was solid.
The release comes three years after SQL Server 2005, which is better than the five-year gap between it and SQL Server 2000. Microsoft (NASDAQ: MSFT) got an earful from customers and promised greater expedience with the release cycles.
"Customers clearly told us [the gap between SQL Server 2000 and SQL Server 2005] was too long," said Ted Kummert, corporate vice president of Microsoft's data platform and storage division. "We heard this and committed from there forward to a 24- to 36-month release cycle and said that that the next release of SQL Server would be available 24 to 36 months from the release of 2005."
Throughout that time, Microsoft offered a community technology preview (CTP), the fancy word for a beta, for customers to download and test. As new features were added, the CTP was updated. This helped not only gather feedback but seed the market so applications are ready on launch day.
Microsoft said that more than 75 large-scale applications are already in production, and more than 1,350 applications are being developed by nearly 1,000 independent software vendors (ISVs) on SQL Server 2008. "Our overall objective was to get a lot of customer feedback and get a lot of apps into production," Kummert said.
Microsoft highlighted a number of big changes: a streamlined upgrade path, integration with services by Oracle and SAP, improved Office 2007 performance, transparent data encryption and a system resource governor.
SQL Server is available in seven different editions, from the free Express version that runs on a desktop computer to the Enterprise edition for mission-critical computing. Pricing remains the same for 2008 editions as it was for the 2005 editions.
One of those editions is new. SQL Server 2008 Web is an Internet-designed database meant specifically for highly available Web applications or hosting environments, according to Microsoft. Hosting partners wanted better features, scalability and pricing for a Web version of SQL, so Microsoft created this version to fit these needs.
Existing customers of SQL Server 2005 will find the pricing most agreeable: free, if they have a Software Assurance support license (and who doesn't?). It is available for a free download from Microsoft's TechNet site. An evaluation copy is planned for release on Thursday.
Chris Aliegro, lead analyst with Directions on Microsoft, said this release of SQL Server isn't quite as monumental a jump as the prior release, but remains important, as SQL Server has become a big business at the company.
"This has clearly become a rock star product for Microsoft," he told InternetNews.com. "It's gone from being interesting to strategic and a billion dollar-plus product, and it's used so widely by other products at Microsoft."
The Web version of the product seems like something aimed at MySQL, which is very popular with Web developers. Aliegro would not say yes or no on that, but added he wouldn't be surprised if that was where Microsoft was aiming.
"Clearly there's a great market opportunity for them to do something like that," he said. "SQL Server has been historically used behind the corporate firewall, not to support Web sites," Aliegro explained. "If there's an opportunity for them to move a product into a lucrative market, they are going to do it, and database-backing Web sites is an opportunity for them."
Originally planned for earlier this year to coincide with Windows Server 2008, SQL Server was delayed to insure the code was solid.
The release comes three years after SQL Server 2005, which is better than the five-year gap between it and SQL Server 2000. Microsoft (NASDAQ: MSFT) got an earful from customers and promised greater expedience with the release cycles.
"Customers clearly told us [the gap between SQL Server 2000 and SQL Server 2005] was too long," said Ted Kummert, corporate vice president of Microsoft's data platform and storage division. "We heard this and committed from there forward to a 24- to 36-month release cycle and said that that the next release of SQL Server would be available 24 to 36 months from the release of 2005."
Throughout that time, Microsoft offered a community technology preview (CTP), the fancy word for a beta, for customers to download and test. As new features were added, the CTP was updated. This helped not only gather feedback but seed the market so applications are ready on launch day.
Microsoft said that more than 75 large-scale applications are already in production, and more than 1,350 applications are being developed by nearly 1,000 independent software vendors (ISVs) on SQL Server 2008. "Our overall objective was to get a lot of customer feedback and get a lot of apps into production," Kummert said.
Microsoft highlighted a number of big changes: a streamlined upgrade path, integration with services by Oracle and SAP, improved Office 2007 performance, transparent data encryption and a system resource governor.
SQL Server is available in seven different editions, from the free Express version that runs on a desktop computer to the Enterprise edition for mission-critical computing. Pricing remains the same for 2008 editions as it was for the 2005 editions.
One of those editions is new. SQL Server 2008 Web is an Internet-designed database meant specifically for highly available Web applications or hosting environments, according to Microsoft. Hosting partners wanted better features, scalability and pricing for a Web version of SQL, so Microsoft created this version to fit these needs.
Existing customers of SQL Server 2005 will find the pricing most agreeable: free, if they have a Software Assurance support license (and who doesn't?). It is available for a free download from Microsoft's TechNet site. An evaluation copy is planned for release on Thursday.
Chris Aliegro, lead analyst with Directions on Microsoft, said this release of SQL Server isn't quite as monumental a jump as the prior release, but remains important, as SQL Server has become a big business at the company.
"This has clearly become a rock star product for Microsoft," he told InternetNews.com. "It's gone from being interesting to strategic and a billion dollar-plus product, and it's used so widely by other products at Microsoft."
The Web version of the product seems like something aimed at MySQL, which is very popular with Web developers. Aliegro would not say yes or no on that, but added he wouldn't be surprised if that was where Microsoft was aiming.
"Clearly there's a great market opportunity for them to do something like that," he said. "SQL Server has been historically used behind the corporate firewall, not to support Web sites," Aliegro explained. "If there's an opportunity for them to move a product into a lucrative market, they are going to do it, and database-backing Web sites is an opportunity for them."
Open source blades?
Rackable also announced it's joining Blade.org, the industry consortium created in 2004 by IBM and Intel (NASDAQ: INTC) that's trying to create open standards for blade servers. Up to now, all of the blades from Hewlett-Packard, Dell and Sun have used their own design standards, meaning you can't install a Sun blade in an HP chassis, for example.
IBM has been an exception to that rule, opening its blades up and making the specs public. The first company to offer a third-party product is Themis, which offers a Sun UltraSPARC-based blade that works in a BladeCenter chassis.
"Openness has been a key part of our strategy for blades from day one," said Tim Dougherty, director of BladeCenter strategy at IBM. "We felt it's the right thing to do to expand the blade market."
The ICE Cube is available in 20- or 40-foot container sizes and can hold up to 1,344 dual-socket blades with quad-core Intel Xeons or 672 quad socket, dual-core AMD Opteron blades.
The BladeCenter chassis slots right into a Rackable ICE Cube, so long as you remember to remove the wheels that are on the chassis. It then uses the ICE Cube power and cooling system for operation and management.
IBM BladeCenter T and HT are available immediately via Rackable Systems and its channel partners
IBM has been an exception to that rule, opening its blades up and making the specs public. The first company to offer a third-party product is Themis, which offers a Sun UltraSPARC-based blade that works in a BladeCenter chassis.
"Openness has been a key part of our strategy for blades from day one," said Tim Dougherty, director of BladeCenter strategy at IBM. "We felt it's the right thing to do to expand the blade market."
The ICE Cube is available in 20- or 40-foot container sizes and can hold up to 1,344 dual-socket blades with quad-core Intel Xeons or 672 quad socket, dual-core AMD Opteron blades.
The BladeCenter chassis slots right into a Rackable ICE Cube, so long as you remember to remove the wheels that are on the chassis. It then uses the ICE Cube power and cooling system for operation and management.
IBM BladeCenter T and HT are available immediately via Rackable Systems and its channel partners
IBM, Rackable Team Up on Blades
IBM and Rackable Systems today announced a joint sales agreement whereby Rackable Systems will offer IBM's BladeCenter systems for installation inside of Rackable's ICE Cube modular datacenters.
The ICE Cube is a standard-size shipping container with all of the racks for an ultradense datacenter installed inside. Customers can fill it with Rackable's line of 1U to 9U racks or they can add a BladeCenter chassis to use IBM blades.
The deal involves the use of IBM BladeCenter T or HT systems, which are NEBS-3/ETSI-compliant, meaning they’re certified for use in telecommunications environments and carrier facilities.
Pund-IT Principle Analyst Charles King said it's a win for both companies. "It gives Rackable a way to jump into the blade solution space without having to spend any money developing it themselves," he said, adding that developing the blades is an expensive process. "And it allows IBM to sell BladeCenter as an OEM product -- not that Rackable will put its logo on it."
Tony Carrozza, senior vice president of sales and marketing for Rackable, said adding the BladeCenter products to the its modular datacenter "puts us in a position to offer customers a broader set of solutions, from cloud computing to enterprise apps, which is what the BladeCenter allows us to do going forward," said
Blades from IBM (NYSE: IBM) are inherently different from a Rackable (NASDAQ: RACK) rack-mounted system in that IBM blades are for more general-purpose computing, whereas Rackable systems are meant for ultradense computing projects.
"We don't have what you would consider to be a true blade product," Carrozza told InternetNews.com. "The BladeCenter is a true blade product with a self-contained fabric, management software, redundancy and resiliency in it," he explained, adding that besides rack servers Rackable typically offers scale-out servers. "It's blade like but doesn't have its own switching fabric built into the chassis," he said.
Plus, Rackable's systems are based on Intel and AMD chips, whereas IBM blades also offer the option of IBM's POWER5 and POWER6 processors.
The ICE Cube is available in 20- or 40-foot container sizes and can hold up to 1,344 dual-socket blades with quad-core Intel Xeons or 672 quad socket, dual-core AMD Opteron blades.
The BladeCenter chassis slots right into a Rackable ICE Cube, so long as you remember to remove the wheels that are on the chassis. It then uses the ICE Cube power and cooling system for operation and management.
IBM BladeCenter T and HT are available immediately via Rackable Systems and its channel partners.
The ICE Cube is a standard-size shipping container with all of the racks for an ultradense datacenter installed inside. Customers can fill it with Rackable's line of 1U to 9U racks or they can add a BladeCenter chassis to use IBM blades.
The deal involves the use of IBM BladeCenter T or HT systems, which are NEBS-3/ETSI-compliant, meaning they’re certified for use in telecommunications environments and carrier facilities.
Pund-IT Principle Analyst Charles King said it's a win for both companies. "It gives Rackable a way to jump into the blade solution space without having to spend any money developing it themselves," he said, adding that developing the blades is an expensive process. "And it allows IBM to sell BladeCenter as an OEM product -- not that Rackable will put its logo on it."
Tony Carrozza, senior vice president of sales and marketing for Rackable, said adding the BladeCenter products to the its modular datacenter "puts us in a position to offer customers a broader set of solutions, from cloud computing to enterprise apps, which is what the BladeCenter allows us to do going forward," said
Blades from IBM (NYSE: IBM) are inherently different from a Rackable (NASDAQ: RACK) rack-mounted system in that IBM blades are for more general-purpose computing, whereas Rackable systems are meant for ultradense computing projects.
"We don't have what you would consider to be a true blade product," Carrozza told InternetNews.com. "The BladeCenter is a true blade product with a self-contained fabric, management software, redundancy and resiliency in it," he explained, adding that besides rack servers Rackable typically offers scale-out servers. "It's blade like but doesn't have its own switching fabric built into the chassis," he said.
Plus, Rackable's systems are based on Intel and AMD chips, whereas IBM blades also offer the option of IBM's POWER5 and POWER6 processors.
The ICE Cube is available in 20- or 40-foot container sizes and can hold up to 1,344 dual-socket blades with quad-core Intel Xeons or 672 quad socket, dual-core AMD Opteron blades.
The BladeCenter chassis slots right into a Rackable ICE Cube, so long as you remember to remove the wheels that are on the chassis. It then uses the ICE Cube power and cooling system for operation and management.
IBM BladeCenter T and HT are available immediately via Rackable Systems and its channel partners.
IBM Follows Through on Data Protection for SMBs
Just three months since it acquired continuous data protection (CDP) startup FilesX, IBM has rebranded its solution as IBM Tivoli Storage Manager (TSM) FastBack and today began pushing its own version to market.
CDP, also known as continuous backup, has been growing in importance as a way for companies to better ensure their files are safeguarded. The process automates file backups each time a change is made, enabling offices to not only restore lost files, but also to reconstruct files as they appeared at any point in the past.
In particular, such offerings are lower-cost offerings aimed at helping small-to-midsized businesses (SMBs) and remote locations -- which typically lack dedicated IT staff -- deal with mounting storage concerns.
"There is more data being distributed and companies want consistent data protection strategies," John Connor, product manager for TSM FastBack
Big Blue said the ease of managing FastBack's disk-based, block-level storage approach makes it suitable for small and remote offices that have little tech support.
"This eliminates issues with backup windows and provides near-instant recovery," Connor said.
FastBack, which IBM acquired earlier this year, also represents a second try on the technology, Lauren Whitehouse, an analyst with Enterprise Strategy Group, told InternetNews.com.
IBM initially had attempted to tweak its TSM top-end product to create a data protection tool tailored for small to midsized firms, called TSM Express. But it wasn't a perfect approach, Whitehouse said.
"They scaled down the comprehensive offering but the TSM Express really didn't meet the need as it was still hefty to manage," she added. Instead, the new FastBack offering "is the next generation of that approach."
Still, for Big Blue, being able to swiftly integrate the product and assimilate it into its sales channel shows its proficiency at technology mergers, Whitehouse said.
"They have the formula for taking in technology they acquire and getting it rolled out in an impressive time frame," she added.
The new offering serves to build on IBM's storage flagship, the venerable Tivoli Storage Manager. IBM also said the new FastBack offering, which has been melded into TSM, complements its CDP software for laptops and desktops, IBM Tivoli Continuous Data Protection for Files.
Along with FastBack, IBM also introduced FastBack for Microsoft Exchange and FastBack for Bare Machine Recovery, designed for system and server migrations. Together, the bundled components are sold as TSM FastBack Center.
While the suite is currently offered only for Windows platform, IBM has plans to develop additional versions, it said.
CDP, also known as continuous backup, has been growing in importance as a way for companies to better ensure their files are safeguarded. The process automates file backups each time a change is made, enabling offices to not only restore lost files, but also to reconstruct files as they appeared at any point in the past.
In particular, such offerings are lower-cost offerings aimed at helping small-to-midsized businesses (SMBs) and remote locations -- which typically lack dedicated IT staff -- deal with mounting storage concerns.
"There is more data being distributed and companies want consistent data protection strategies," John Connor, product manager for TSM FastBack
Big Blue said the ease of managing FastBack's disk-based, block-level storage approach makes it suitable for small and remote offices that have little tech support.
"This eliminates issues with backup windows and provides near-instant recovery," Connor said.
FastBack, which IBM acquired earlier this year, also represents a second try on the technology, Lauren Whitehouse, an analyst with Enterprise Strategy Group, told InternetNews.com.
IBM initially had attempted to tweak its TSM top-end product to create a data protection tool tailored for small to midsized firms, called TSM Express. But it wasn't a perfect approach, Whitehouse said.
"They scaled down the comprehensive offering but the TSM Express really didn't meet the need as it was still hefty to manage," she added. Instead, the new FastBack offering "is the next generation of that approach."
Still, for Big Blue, being able to swiftly integrate the product and assimilate it into its sales channel shows its proficiency at technology mergers, Whitehouse said.
"They have the formula for taking in technology they acquire and getting it rolled out in an impressive time frame," she added.
The new offering serves to build on IBM's storage flagship, the venerable Tivoli Storage Manager. IBM also said the new FastBack offering, which has been melded into TSM, complements its CDP software for laptops and desktops, IBM Tivoli Continuous Data Protection for Files.
Along with FastBack, IBM also introduced FastBack for Microsoft Exchange and FastBack for Bare Machine Recovery, designed for system and server migrations. Together, the bundled components are sold as TSM FastBack Center.
While the suite is currently offered only for Windows platform, IBM has plans to develop additional versions, it said.
IBM Thinking Green for N.C. Datacenter
IBM today announced that it will build an advanced datacenter in its Research Triangle Park, Raleigh, N.C., facility for $360 million. This will be the first datacenter to be built with the computer giant's New Enterprise Datacenter design principles.
The New Enterprise Datacenter platform is a fusion of Google's Webcentric cloud approach and the MySpace approach, with an emphasis on data-intensive parallel programming.
The datacenter will be one part of a hub for IBM's (NYSE: IBM) computing infrastructure in the cloud that clients will be able to access anytime from anywhere. The other part of the hub is IBM's datacenter in Tokyo, which is also being unveiled today.
Data-intensive parallel computing "turned out to be a very key element of cloud computing applications we've seen deployed by Facebook, Yahoo (NASDAQ: YHOO), Google (NASDAQ: GOOG) and some folks in the telecommunications industry we've had discussions with," Dennis Quan, director of development, autonomic computing, with the IBM software group, told InternetNews.com.
The technologies for both datacenters were shaped by work done through IBM's partnership with Google. The two teamed up in October to create three datacenters for academic computing using data-intensive parallel programming.
IBM's work with Google "gave us a lot of knowledge and research to figure out the best way to bring out the cloud computing platform," Quan explained. "The manifestation of that is our Unified Datacenter Architecture."
According to Quan, this architecture consists of IBM's systems technologies -- virtualized networks, storage, compute resources using Xen and VMware on x86 boxes and native virtualization capabilities on power systems and the mainframe.
It also includes using Tivoli software "to drive provisioning and monitoring of systems in the datacenter, and advanced capabilities like doing chargeback and ensuring high levels of availability and storage management," Quan added.
Befriending the environment
The Raleigh, N.C., datacenter will leverage green computing principles. It will use many of the technologies from IBM's Project Big Green initiative to slash energy consumption, and use the latest water-cooled and air-cooled equipment.
IBM will make heavy use of virtualization technology in the datacenter. The company plans to adopt a modular construction technology to minimize cost and environmental impact.
After the first 60,000 feet are built, additional space will be added in equal-size modules on demand. The center will be able to support 2.5 to three times as many clients as a traditional datacenter of comparable size, IBM said.
Work on the Raleigh, N.C., datacenter will begin soon, and it's expected to be completed by next year. It will be "the biggest of the cloud centers we've launched," Quan said.
Since March, IBM has launched cloud computing centers in Dublin, Ireland; Beijing, China; and Johannesburg, South Africa. The Tokyo and Raleigh, N.C., datacenters will be the vendor's eighth and ninth datacenters, respectively.
Over time, Quan expects enterprises to increasingly leverage cloud computing at its datacenters. "We're going to see a lot of apps come about based on SOA (define) that will allow applications on various clouds to talk to each other and be very functional and be managed in a centralized fashion," he said. SOA is service-oriented architecture.
Charles King, principal analyst at Pund-IT, said IBM is making "a very large investment in what's still a very interesting, but from a commercial standpoint, an emerging technology" with its move into cloud computing. However, the datacenters "point to IBM's strength in datacenter design and hardware strength."
King contrasted IBM's moves to the joint initiative announced recently by Hewlett-Packard (NYSE: HP), Intel (NASDAQ: INTC) and Yahoo to set up cloud computing labs. "They're working on research testbeds for cloud computing, whereas IBM is moving forward with what will be commercial datacenters,
Other cloud computing initiatives
HP has another cloud computing initiative it launched in May. This will help enterprises build their own minicloud datacenters, complete with products and services for building out the center.
Cloud computing is fast becoming a growth area, and companies are flocking to this sector thick and fast.
In February, storage giant EMC bought Seattle-based startup Pi, using it as the nexus of its newly set-up Cloud Infrastructure and Services Division. Pi founder Paul Maritz, a 14-year veteran of Microsoft (NASDAQ: MSFT), became head of that division.
Maritz is now CEO of virtualization giant VMware, an EMC (NYSE: EMC) company, and sits on its board of directors. He took the post after VMware co-founder and CEO Diane Greene left abruptly.
In June, Q-layer announced the concept of a virtual private datacenter. This would let companies support on-demand cloud computing through virtual datacenters, according to a report by Enterprise Strategy Group analyst Mark Bowker. It would also let them build their own virtual environments, using a credit-based chargeback system to track resource allocation and utilization.
Also in June, Red Hat (NYSE: RHT) made its JBoss Enterprise Platform available on Amazon's (NASDAQ: AMZN) EC2. This lets users build, deploy and host enterprise Java applications and services in the cloud.
And earlier this month, RightScale and GigaSpaces teamed up to let enterprises deploy and scale data- and transaction-intensive applications on EC2, and manage them. And 3Tera lets customers run applications in their own datacenter.
The New Enterprise Datacenter platform is a fusion of Google's Webcentric cloud approach and the MySpace approach, with an emphasis on data-intensive parallel programming.
The datacenter will be one part of a hub for IBM's (NYSE: IBM) computing infrastructure in the cloud that clients will be able to access anytime from anywhere. The other part of the hub is IBM's datacenter in Tokyo, which is also being unveiled today.
Data-intensive parallel computing "turned out to be a very key element of cloud computing applications we've seen deployed by Facebook, Yahoo (NASDAQ: YHOO), Google (NASDAQ: GOOG) and some folks in the telecommunications industry we've had discussions with," Dennis Quan, director of development, autonomic computing, with the IBM software group, told InternetNews.com.
The technologies for both datacenters were shaped by work done through IBM's partnership with Google. The two teamed up in October to create three datacenters for academic computing using data-intensive parallel programming.
IBM's work with Google "gave us a lot of knowledge and research to figure out the best way to bring out the cloud computing platform," Quan explained. "The manifestation of that is our Unified Datacenter Architecture."
According to Quan, this architecture consists of IBM's systems technologies -- virtualized networks, storage, compute resources using Xen and VMware on x86 boxes and native virtualization capabilities on power systems and the mainframe.
It also includes using Tivoli software "to drive provisioning and monitoring of systems in the datacenter, and advanced capabilities like doing chargeback and ensuring high levels of availability and storage management," Quan added.
Befriending the environment
The Raleigh, N.C., datacenter will leverage green computing principles. It will use many of the technologies from IBM's Project Big Green initiative to slash energy consumption, and use the latest water-cooled and air-cooled equipment.
IBM will make heavy use of virtualization technology in the datacenter. The company plans to adopt a modular construction technology to minimize cost and environmental impact.
After the first 60,000 feet are built, additional space will be added in equal-size modules on demand. The center will be able to support 2.5 to three times as many clients as a traditional datacenter of comparable size, IBM said.
Work on the Raleigh, N.C., datacenter will begin soon, and it's expected to be completed by next year. It will be "the biggest of the cloud centers we've launched," Quan said.
Since March, IBM has launched cloud computing centers in Dublin, Ireland; Beijing, China; and Johannesburg, South Africa. The Tokyo and Raleigh, N.C., datacenters will be the vendor's eighth and ninth datacenters, respectively.
Over time, Quan expects enterprises to increasingly leverage cloud computing at its datacenters. "We're going to see a lot of apps come about based on SOA (define) that will allow applications on various clouds to talk to each other and be very functional and be managed in a centralized fashion," he said. SOA is service-oriented architecture.
Charles King, principal analyst at Pund-IT, said IBM is making "a very large investment in what's still a very interesting, but from a commercial standpoint, an emerging technology" with its move into cloud computing. However, the datacenters "point to IBM's strength in datacenter design and hardware strength."
King contrasted IBM's moves to the joint initiative announced recently by Hewlett-Packard (NYSE: HP), Intel (NASDAQ: INTC) and Yahoo to set up cloud computing labs. "They're working on research testbeds for cloud computing, whereas IBM is moving forward with what will be commercial datacenters,
Other cloud computing initiatives
HP has another cloud computing initiative it launched in May. This will help enterprises build their own minicloud datacenters, complete with products and services for building out the center.
Cloud computing is fast becoming a growth area, and companies are flocking to this sector thick and fast.
In February, storage giant EMC bought Seattle-based startup Pi, using it as the nexus of its newly set-up Cloud Infrastructure and Services Division. Pi founder Paul Maritz, a 14-year veteran of Microsoft (NASDAQ: MSFT), became head of that division.
Maritz is now CEO of virtualization giant VMware, an EMC (NYSE: EMC) company, and sits on its board of directors. He took the post after VMware co-founder and CEO Diane Greene left abruptly.
In June, Q-layer announced the concept of a virtual private datacenter. This would let companies support on-demand cloud computing through virtual datacenters, according to a report by Enterprise Strategy Group analyst Mark Bowker. It would also let them build their own virtual environments, using a credit-based chargeback system to track resource allocation and utilization.
Also in June, Red Hat (NYSE: RHT) made its JBoss Enterprise Platform available on Amazon's (NASDAQ: AMZN) EC2. This lets users build, deploy and host enterprise Java applications and services in the cloud.
And earlier this month, RightScale and GigaSpaces teamed up to let enterprises deploy and scale data- and transaction-intensive applications on EC2, and manage them. And 3Tera lets customers run applications in their own datacenter.
IBM Unveils Proactive E-Discovery Solution
IBM has unveiled eDiscovery Manager, a software product that lets enterprises manage electronically stored information so that they can retrieve it easily when a legal challenge requires e-discovery. The term e-discovery stands for 'electronic discovery' and refers to discovery in civil litigation dealing with information in electronic form.
Enterprises are being forced to adopt e-discovery solutions because of amendments in December to the Federal Rules of Civil Procedure (FRCP) (define) that require companies to preserve and produce electronically stored information when facing litigation. FRCP is the U.S. federal district court procedures for civil suits.
Part of the vendor's Enterprise Content Management (ECM) (define) suite of products, eDiscovery Manager integrates with IBM's auto-classification and records management technology, and the vendor's content-centric business process management (BPM) (define) capabilities.
The eDiscovery Manager uses IBM's e-mail archiving solutions, leverages the vendor's ECM repositories, and supports an easy-to-use interface. It "controls information at its source when it is created," Aaron Brown, program director of IBM Content Discovery told InternetNews.com. The eDiscovery Manager works with IBM's Classification Module, and content management repository for this.
"We let you handle capture, retention, archiving and content management to manage your content proactively because the information has already been retained, classified and managed," Brown said. Many other e-discovery vendors offer point solutions or outsourced third-party solutions that companies put in place when litigation comes in, according to Brown.
While these point and third-party outsourced approaches do work, "they are expensive, because organizations often have to expand them to deal with other cases, and typically these solutions don't address the bigger problem of information retention," Brown said. "We want e-discovery to be part of the basic daily process rather than having you run around putting out fires."
IBM (NYSE: IBM) is one of the few vendors that can offer an enterprise-class solution for eDiscovery, Rob Enderle, principal analyst at the Enderle Group, told InternetNews.com "Nobody has the breadth anymore to cover all databases, all repositories; it requires a very large set of capabilities and IBM Global Services and EDS are among the few entities that folks love to help solve that kind of complex problem."
EDS is now a part of Hewlett-Packard (NYSE: HPQ), which bought in May for $13.9 billion.
Together with other products in IBM's ECM suite, eDiscovery Manager lets enterprises sort, classify and archive information for easy retrieval.
The inner workings
IBM eDiscovery Manager is an integral part of the IBM Compliance Warehouse for Legal Control, a combination of software, hardware and services that let enterprises achieve, sustain and prove compliance with multiple legal and compliance mandates
It also supports IBM's broader Information Governance strategy, which helps clients define, enforce and monitor policies related to the control and quality of information. IBM eDiscovery Manager runs on Windows and AIX, IBM's version of Unix. IBM is moving AIX to open source.
When lawyers need information relating to a case, they provide the parameters such as the subject, a range of dates, and keywords to IT, which conducts a search using eDiscovery Manager's search-based interface. The product collects relevant information and puts a litigation hold on it, which marks that information as undeletable.
"A very important part of the eDiscovery process is that the tools you use must respect things like deletion policies and keep accurate logs of access and searches," Brown said. These functions have been built into eDiscovery Manager, he added.
Once the search is complete, the IT department checks the information gathered and exports the result to the legal team, Brown said. "This is not the end application for the legal team, we're just focused on proactive management and collecting and identifying the information
Enterprises are being forced to adopt e-discovery solutions because of amendments in December to the Federal Rules of Civil Procedure (FRCP) (define) that require companies to preserve and produce electronically stored information when facing litigation. FRCP is the U.S. federal district court procedures for civil suits.
Part of the vendor's Enterprise Content Management (ECM) (define) suite of products, eDiscovery Manager integrates with IBM's auto-classification and records management technology, and the vendor's content-centric business process management (BPM) (define) capabilities.
The eDiscovery Manager uses IBM's e-mail archiving solutions, leverages the vendor's ECM repositories, and supports an easy-to-use interface. It "controls information at its source when it is created," Aaron Brown, program director of IBM Content Discovery told InternetNews.com. The eDiscovery Manager works with IBM's Classification Module, and content management repository for this.
"We let you handle capture, retention, archiving and content management to manage your content proactively because the information has already been retained, classified and managed," Brown said. Many other e-discovery vendors offer point solutions or outsourced third-party solutions that companies put in place when litigation comes in, according to Brown.
While these point and third-party outsourced approaches do work, "they are expensive, because organizations often have to expand them to deal with other cases, and typically these solutions don't address the bigger problem of information retention," Brown said. "We want e-discovery to be part of the basic daily process rather than having you run around putting out fires."
IBM (NYSE: IBM) is one of the few vendors that can offer an enterprise-class solution for eDiscovery, Rob Enderle, principal analyst at the Enderle Group, told InternetNews.com "Nobody has the breadth anymore to cover all databases, all repositories; it requires a very large set of capabilities and IBM Global Services and EDS are among the few entities that folks love to help solve that kind of complex problem."
EDS is now a part of Hewlett-Packard (NYSE: HPQ), which bought in May for $13.9 billion.
Together with other products in IBM's ECM suite, eDiscovery Manager lets enterprises sort, classify and archive information for easy retrieval.
The inner workings
IBM eDiscovery Manager is an integral part of the IBM Compliance Warehouse for Legal Control, a combination of software, hardware and services that let enterprises achieve, sustain and prove compliance with multiple legal and compliance mandates
It also supports IBM's broader Information Governance strategy, which helps clients define, enforce and monitor policies related to the control and quality of information. IBM eDiscovery Manager runs on Windows and AIX, IBM's version of Unix. IBM is moving AIX to open source.
When lawyers need information relating to a case, they provide the parameters such as the subject, a range of dates, and keywords to IT, which conducts a search using eDiscovery Manager's search-based interface. The product collects relevant information and puts a litigation hold on it, which marks that information as undeletable.
"A very important part of the eDiscovery process is that the tools you use must respect things like deletion policies and keep accurate logs of access and searches," Brown said. These functions have been built into eDiscovery Manager, he added.
Once the search is complete, the IT department checks the information gathered and exports the result to the legal team, Brown said. "This is not the end application for the legal team, we're just focused on proactive management and collecting and identifying the information
Subscribe to:
Posts (Atom)
