Friday, March 20, 2009

If you’re an enterprise, developer or economist, IBM is not the right buyer for Sun

From the perspective of IT users, developer communities and global industry as a whole, IBM may be the worst place for beleaguered Sun Microsystems to land.

Sure a merger as is rumored is good -- but not urgently or obviously so -- for IBM. Big Blue gains modest improvement in share of some servers, mostly Unix-based. It would actually gain just enough share of high-end servers to justly draw anti-trust scrutiny nearly worldwide.

Yet these types of servers are not today's growth engines for IT vendors, they are the blunt trailing edge. Users have been dumping them in droves, with their sights set on far lower-cost alternatives and newer utility models of deployment and payment. IBM may want the next generation of data centers to be built of mainframes, but not too many others do.

In any event, server hardware is not a meaningful differentiator in today’s IT markets. Sun, if anyone, has proven that. IBM to claim it as the rationale for the buyout is fishy. A lot of other analysts are holding their noses too. UPDATE: Good analysis from Redmonk's Stephen O'Grady.

The rumored IBM-Sun deal for $6.4 billion is incremental improvement for IBM on several fronts: open source software (low earnings), tape storage (modest albeit dependable revenue), Java (already mostly open), engineering talent (easier to get these days given Sun layoffs), new intellectual property (targeted by design by Sun on undercutting IBM's cash cows). In short, there are no obvious game changers or compelling synergies in IBM buying Sun other than setting the sun on Sun.

I initially thought the rumored deal, which drove up Sun's stock, JAVA, by nearly 80 percent on rumor day one, didn't make sense. But it does make sense. Unfortunately it only makes sense for IBM in a fairly ugly way. As Tom Foremski said, it smacks of a spoiler role.

If IBM, would you spend what may end up being $4 billion in actual cost to slow or stifle the deterioration of a $100 billion data center market, and, at the same time, take the means of accelerating the move to cloud computing off the table from your competitors? As Mister Rogers would say, "Sure, sure you would."

Most likely, though the denials are in the works, IBM will plunder and snuff, plunder and snuff its way across the Sun portfolio -- from large account to large account, developer community to developer community, employee project to project. The tidy market share and technology gems will be absorbed quietly, the rest canceled or allowed to wither on the vine.

Certain open source communities and projects that Sun has fostered will be cultivated, or not. IBM is the very best at knowing how to play the open source cards, and that does not mean playing them all.

Listen, this would be a VERY different acquisition than any IBM has done in recent memory. It’s really about taking a major competitor out when they are down. It’s bold and aggressive, and it’s ignoble. But these are hard times and many people are distracted.

The deal is not good for Sun and it's customers (unless they already decided to move from being a Sun shop to an IBM shop), and may put in jeopardy the momentum of open source use up into middleware, SOA, databases and cloud infrastructure. That’s because, even at the price of $6.4 billion (twice Sun's market value before the deal talk), IBM will gain far more from the deal over the long term by eradicating Sun than by joining Sun's vector.

This deal is all about control. Control of Java, of markets, developers, cost of IT -- even about the very pace of change across the industry. For much of it's history IBM has had its hand on the tiller of the IT progression. It's was a comfortable position except for an historically exceptional past 17 years for IBM. It's time to get back in the saddle.

Clearly, Sun has little choice in the matter, other than to jockey for the best price and perhaps some near-term concessions for its employees. It's freaking yard sale. Sun is being run by -- gasp -- investment bankers. Here's a rare bonus bonanza in a M&A desert, for sure.

But let's be clear, this is no merger of partners or equals. This is assimilation. It’s Borg-like, and resistance may be futile. It is important to know when you're being assimilated, however.

Scott McNealy, Sun’s chairman, former CEO and co-founder, famously called the 2001 proposed merger of HP and Compaq a collision between two "garbage trucks." Well, IBM’s proposed/rumored purchase of Sun is equivalent to a garbage truck being airlifted out of sight and over the horizon by a C-17 cargo transport plane. Just open the door and drive it in. The plane was probably designed on Sun hardware, too. [Disclosure: HP is a sponsor of BriefingsDirect podcasts.]

Sun’s fate has been shaky for a long time now. The reasons are fodder for Harvard case studies.

But what of the general good of enterprise IT departments, of communities of idealistic developers, or of open and robust competition in the new age of cloud computing? In the new age, incidentally, you may no longer need an army of consultants and C-17 full of hardware and software at each and every enterprise. As Nick Carr correctly points out, this changes everything. That kind of change may not be what IBM has in mind.

It’s not easy resting having IBM in control of a vast portions of the open source future, and the legacy installed past. Linux and Apache Web servers might have made sense for IBM, but do open source cloud databases, middleware, SOA, and the next generations of on- and off-premises utility and virtualization fabric infrastructure?

IBM today is making the lion's share of its earnings from the software and services that run yesterday's data centers. Even the professional services around the newer cloud models (and subscription fees of actual, not low-utilization, use) does not make up for lost software license revenues. In many ways, cloud is more a threat than an opportunity to Big Blue. It ultimately means lower revenues, lower margins, less control, and feisty competitors that make money from ads and productivity, not sales and service.

Cloud models will take a long time to become common and mainstream, but any sense of inevitability must make IBM (and others) nervous. Controlling the pace of the change is essential.

The hastening shift to virtualization, application modernization, SaaS, mobile, cloud, and increased use of open source for legacy infrastructure could seriously disrupt the business models of IBM, HP, Cisco, Microsoft, Oracle and others. Moving from legacy-and-license to cloud-and-subscription (on OSS or commercial code) poses a huge risk to IBM, especially if it happens fast -- something this unexpected economic crisis could accelerate.

Enterprises could soon gain the equivalent of the powerful and efficient IT engines that run a Google or Amazon, either for itself, or rented off the wire, or both. IBM probably won't have 60 percent of the cloud services market in five years like it does the high-end Unix market (if it gets Sun). In fact, what has happened to Sun in terms of disruption may be a harbinger of could happen to IBM during the next red-shift in the market.

Sun should have gotten to these compelling cloud values first, made a business of it before Amazon. Sun was on the way, had the vision, but ran out of time and out of gas.

Sun has let a lot of us down by letting it come to this. The private equity firms that control Sun now don't give a crap about open source, or innovation, clouds or whether the network is the computer, or my dog's pajamas are the computer. They need to get their money back ASAP.

As a result, they and Sun could well be handing over to IBM the very keys to being able to time the market to IBM's strategic needs above all else. All for $6.4 billion in cash, minus the profits from chopping off Sun's remaining limbs and keeping the ones that make a good Borg fit.

There should be a better outcome. Should the deal emerge, regulators should insist what IBM itself called for more than 10 years ago. Something as important as Java and other critical open software specifications (OpenSolaris?) should be in the control and ownership of a neutral standards body, not in the control of the global market dominant legacy vendor.

It’s sort of like letting General Motors decide when to build the next generation of fuel efficient and alternative energy cars. And we know how that worked out.

IBM has the deep pockets now to buy strategic advantage during an economic crisis that helps it in coming years. It's during this coming period when the cloud vision begins to stick, when the madness of how enterprise IT has evolved in cost and complexity is shaken off for something much better, faster and cheaper.

And that’s what IT has always been about.

Wednesday, March 18, 2009

Greenplum aims to eliminate massive data load 'choke points' with Scatter/Gather technology

Greenplum has taken massively parallel processing (MPP) of data to the next level with the introduction this week of its "MPP Scatter/Gather Streaming" (SG Streaming) technology, which manages the flow of data into all nodes of the database, eliminating the traditional bottlenecks with massive data loading.

The San Mateo, Calif. company, which provides large-scale analytics and data warehousing, says SG Streaming has allowed customers to achieve production-loading speeds of over four terabytes per hour with negligible impacts on concurrent database operations. [Disclosure: Greenplum is a sponsor of BriefingsDirect podcasts.]

Under the "parallel everywhere" approach to loading data flows from one or more source systems to every node of the database without any sequential choke points. This differs from traditional “bulk loading” technologies, used by most mainstream database and parallel-processing appliance vendors that push data from a single source, often over a single or small number of parallel channels, and result in fundamental bottlenecks and ever-increasing load times.

The new technology "scatters" data from all source systems across hundreds or thousands of parallel streams that simultaneously flow to all nodes of the database. Performance scales with the number of nodes, and the technology supports both large batch and continuous near-real-time loading patterns with negligible impact on concurrent database operations.

Data can be transformed and processed in-flight, utilizing all nodes of the database in parallel, for extremely high-performance extract-load-transform (ELT) and extract-transform-load-transform (ETLT) loading pipelines. Final 'gathering' and storage of data to disk takes place on all nodes simultaneously, with data automatically partitioned across nodes and optionally compressed.

It was just six months ago that Greenplum publicly unveiled how it wrapped MapReduce approaches into the newest version of its data solution. That advance allowed users to combine SQL queries and MapReduce programs into unified tasks executed in parallel across thousands of cores.

Active Endpoints aims at greater process design and implementation productivity with ActiveVOS enhancements

Active Endpoints, maker of the ActiveVOS visual orchestration system, has kicked things up a notch with the recent release of ActiveVos 6.1, which incorporates new features and functions designed to make developers more productive.

The latest offering from the Waltham, Mass. company provides what amounts to shrink-wrapped service-oriented architecture (SOA) and provides business process management (BPM) automation, while adhering to business process execution language (BPEL) standards. [Disclosure: Active Endpoints is a sponsor of BriefingsDirect podcasts.]

There's an Active Endpoints podcast on the solution, and a new white paper on SOA implications of the process efficiencies from Dave Linthicum. We also recently did an Analyst Insights podcast on recent BPEL4People work.

Following close on the heels of version 6.0, which debuted in September, and 6.0.2, which made its appearance in December, the newest ActiveVOS offering brings features aimed at smoothing the way for developers. For example, a new tool, the "participant's view," eliminates the need for developers to manually code complex programming constructs like BPEL partner links and BPEL partner link types that are needed to define how services are to be used in a BPM application.

Another major enhancement is "process rewind." At design time, no BPM application can anticipate all of the operational issues and error handling that will be required. Process rewind gives developers the ability to rewind a process to a specific activity and redo the work without having to invoke any of the built-in compensation logic. This allows certain steps of the process need to be “redone” without impacting work already performed.

Among the other improvements:
  • Any-order development, which presents services details as graphical tables into which details can be entered at any time. This is in contrast to earlier systems in which developers needed to know the details in advance.

  • Automatic development, which eases the tasks for developers new to SOA-based BPM. Version 6.1 automatically understands “private” versus “public” web services description language (WSDL) files and creates the required WSDLs in both a standards-compliant mode and a human-understandable format.

  • Improved data handling, which allows developers to visually specify what data is needed in each activity and guides the developer through XPath and XQuery statement generation. The BPEL standard separates assignment of data to activities from the invocation of those activities. While the technical reasons for this are clear to experienced developers, for new developers this can be an impediment.
More information on new features and functions is available on the Active Endpoints "What's New" page on their Web site.

ActiveVOS is available as a perpetual license. In an internal development environment, the price is $5,000 per CPU socket. In a deployment environment, the price is $12,000 per CPU socket when the deployment environment licenses are ordered with a first-time purchase of internal development environment licenses. Annual support and maintenance is 20 percent of total license fees.

Panda Security strengthens SaaS-based PC virus protection solution for SMBs

As the whirlwind of economic pressures and heightened concerns for security push small and medium -sized businesses (SMB) toward software-as-a-service (SaaS) solutions, Panda Security has delivered added functionality to the cause with Managed Office Protection (MOP) 5.03.

Panda, with North American operations in Glendale, CA, allows individual companies as well as value added resellers (VARs) to deploy and extend its hosted security services, which originally launched in May 2008. Panda says its solution can be more than 50 percent more efficient than traditional endpoint security software.

I expect that SMBs will be more likely to seek a full package of PC support services via third parties. Those third parties will want to deliver help desk, software management, patch management and -- now -- security as a full service, cloud-based offering.

By adding the Web-based Panda SaaS security benefits, branded under the third parties, the hassle and cost of managing each desktop on premises drops significantly. And it allows the SMBs to get closer to their goal of no IT department, or at least a majority of IT support gained as a service.

Enhancements to Panda's MOP, include:
  • Optimized management of end devices through a new Web-based management console that allows administrators to resolve deployment challenges from one centralized dashboard from on any computer with an Internet connection.

  • Increased reporting flexibility that allows administrators to select from an expanded set of security reports, including executive, activity and detection reports.

  • Easier software deployment, which allows IT managers to leverage automatic uninstallers along with unique MAC addresses, facilitating personalized security settings for each end-device.

  • Simplified computer management that allows offline handling of exported files.

  • Improved client network status control, which allows VARs providing security services to SMB clients to have remote access via the service provider administration console, where they can centrally manage any update on every device in the client network.
The appeal of using a SaaS solution for security is that for cash-strapped companies it eliminates high startup and capital equipment costs, as well as the necessity to hire and train personnel to run the application. Also, it facilitates updates and patches, allowing them to be deployed quickly and easily.

The channel and PC support third parties gain a more complete package of services, while letting their partner, in this case Panda, pick up the security and on-going threats response requirements.

Another benefit comes from today's highly mobile workforce. Administrators are increasingly concerned with managing laptops belonging to traveling employees. A SaaS-based device support solution allows administrators to monitor and configure anti-malware software no matter what the employee's location.

In a recent study, Panda Security compared its SaaS product to three different traditional security products. The study found that using a SaaS product could be more than 50 percent less expensive over a two-year period than using the traditional products, when you consider staffing costs, capital expenditures, and deployment costs.

Panda MOP is available immediately in licenses sold by the seat in one- to three-year subscription packages. More information is available from www.pandasecurity.com.

IBM buying Sun Microsystems makes no sense, it's a red herring

Someone has floated a trial balloon, through a leak to the Wall Street Journal, that IBM is in "talks" to buy Sun Microsystems for $6.5 billion. The only party that would leak this information is Sun itself, and it smacks of desperation in trying to thwart an unwanted acquisition, or to positively impact another deal that Sun is weak in.

If IBM wanted to buy Sun it would have done so years ago, at least on the merits of synergy and technology. If IBM wanted to buy Sun simply to trash the company, plunder the spoils and do it on the cheap -- the time for that was last fall.

So more likely, given that Sun has reportedly been shopping itself around (nice severance packages for the top brass, no doubt), is that Sun has been too successful at selling itself -- just to the wrong party at too low of a price. This may even be in the form of a chop shop takeover. The only thing holding up a hostile takeover of Sun to sell for spare parts over the past six months was the credit crunch, and the fact that private equity firms have had some distractions.

By buying Sun IBM gains little other than some intellectual property and mySQL. IBM could have bought mySQL or open sourced DB2 or a subset of DB2 any time, if it wanted to go that route. IBM has basically already played its open source hand, which it did masterfully at just the right time. Sun, on the other hand, played (or forced) its open source hand poorly, and at the wrong time. What's the value to Sun for having "gone open source"? Zip. Owning Java is not a business model, or not enough of one to help Sun meaningfully.

So, does IBM need chip architectures from Sun? Nope, has their own. Access to markets from Sun's long-underperforming sales force? Nope. Unix? IBM has one. Linux? IBM was there first. Engineering skills? Nope. Storage technology? Nope. Head-start on cloud implementations? Nope. Java license access or synergy? Nope, too late. Sun's deep and wide professional services presence worldwide? Nope. Ha!

Let's see ... hardware, software, technology, sales, cloud, labor, market reach ... none makes sense for IBM to buy Sun -- at any price. IBM does just fine by continuing to watch the sun set on Sun. Same for Oracle, SAP, Microsoft, HP.

With due respect to Larry Dignan on ZDNet, none of his reasons add up in dollars and cents. No way. Sun has fallen too far over the years for these rationales to stand up.

Only in playing some offense via data center product consolidation against HP and Dell would buying Sun help IBM. And the math doesn't add up there. The cost of getting Sun is more than the benefits of taking money from enterprise accounts from others. [Disclosure: HP is a sponsor of BriefingsDirect podcasts.]

The cost of Sun is not cheap, or at least not cheap like a free puppy. Taking over Sun for technology and market spoils ignores the long-term losses to be absorbed, the decimated workforce, the fact that Cisco will now eat Sun's lunch as have the other server makers for more than five years.

So who might by Sun on the cheap, before Sun's next financial report to Wall Street? Cisco, Dell, EMC, Red Hat. That's about it for vendors. And it would be a big risk for them, unless the price tag were cheap, cheap, cheap. Anything under $4 billion might make sense. Might.

Other buyers could come in the form of carriers, cloud providers or other infrastructure service provider types. This is a stretch, because even cheap Sun would come with a lot of baggage for their needs. Another scenario is a multi-party deal, of breaking up Sun among several different kinds of firms. This also is hugely risky.

So my theory -- and it's just a guess -- is that today's trial balloon on an IBM deal is a last-ditch effort by Sun to find, solidify, or up the price on some other acquisition or exit strategy by Sun. The risk of such market shenanigans only underscores the depths of Sun's malaise. The management at Sun probably sees its valuation sinking yet gain to below tangible assets and cash value when it releases it's next quarterly performance results. ... Soon.

The economic crisis has come at a worst time for Sun than just about any other larger IT vendor. Sun, no matter what happens, will go for a fire sale deal -- not a deal of strength among healthy synergistic partners. No way.