Google’s ActiveX and Sandboxing
December 11, 2008
Google uses unusual words to describe its inventions. If you poke around, you will find that Google has an innovation called “containers”. PCWorld’s Robert McMillan wrote about a Google technology called Native Client. There were some similarities between ActiveX and Native Client but Native Client echoed the container method. The lineage of some Google inventions is murky, but where you read Mr. McMillan’s story here, note this comment:
Called Native Client, the software was released under an open-source software license by Google engineers on Monday. It’s still in an early stage of development, but Google says it could eventually help Web developers create Web programs that would run more quickly and feel more like real desktop applications.
In my opinion, Google is signaling that it wants to provide developers with ways to “hook” more sophisticated applications into Google’s infrastructure. Unlike ActiveX, Google seems to be moving slowly so that certain security issues will be less likely to plague Native Client. Do you see influences of “sandboxing”? I do. I also found it interesting that Mr. McMillan reported that Native Client works in most browsers with the notable exception of Microsoft’s Internet Explorer. Speed is a good thing, and I think Google is blending speed and security which I find another indication of the firm’s maturation. You can find Google’s original Web log post here. Track other Google developer announcements here.
Stephen Arnold, December 11, 2008
Video Horserace
December 10, 2008
Many Web log wizards are chasing the story about Google’s adding magazine content to its burgeoning commercial database killing service. Old news from my point of view. The GOOG is becoming the go-to service for research with words. In fact, it is game over for Thomson Reuters, Reed Elsevier, Wolters Kluwer, Ebsco, and others in this professional content sector. Management at these companies can whip up Excel models to prove me wrong, but Google has demographics and infrastructure on its side. Besides I describe the trajectory of word-centric content in my forthcoming Google and Publishing study.
The real action is in video. Anyone under the age of 17 can explain what’s happening in information. Words are okay, but the future is the rich media experience where words are amplified by music. Do you have a soundtrack for your life? My neighbors’ kids do. Do you make pix and vids on your mobile phone and use these to perform communication functions for which I need pencil and paper? Check out the demographics, and you don’t need the flailing New York Times to remind you of trouble when reporting that Harcourt will not publish books for a while.
Consider the comScore video results table here. The handicappers look at the data, which are probably generally on track but off the mark in absolute values, and see that Google is at the top of the table. Google appears to have about 40 percent of the online video traffic measured and analyzed by comScore. So 60 percent of the video traffic is “in play”; that is, other companies can enter the video ballgame and have some room to maneuver. Look at the number two player, Fox Interactive Media. If comScore data are reasonably accurate, Fox has a chokehold on four percent of the videos viewed market. One of the largest media companies in the world has captured four percent of the market and lags Google’s YouTube by 36 percent. The rest of the field perform less well. Hulu.com, the darling of the old TV kingpins, is in the race. Maybe Hulu.com like a marathon runner getting a marvelous second wind can close the gap between Hulu and Google, but I think I will give Google the advantage for now.
Who cares? The action is text, right?
Wrong, wrong, wrong. YouTube.com could be a major cost sinkhole for Google. If video is expensive for the GOOG, how much of a dent in the bank accounts will video make at outfits like Fox, NBC, and others in the comScore table. Google, for now, seems will to spend to support YouTube.com. As the credit mistral whips through old media, a willingness to spend may winnow the companies in the comScore league table.
Demographics and time, therefore, may give Google an advantage. As pundits gnash their teeth over Google’s overt moves into commercial textual information, Google management is implementing tactics designed to bleed rich media companies, thus weakening them.
Just as the book publishers and other print gurus rolled over into a position of submission to Googzilla, the same fate awaits rich media. Google Books’ growth is old news. The real action is in rich media. The comScore table makes clear to me that the GOOG is poised to destablize more 20th century giants with its 21st century business model. Now tell me why I am incorrect. Facts, please. Catcalls make the geese honk.
Stephen Arnold, December 9, 2008
Google: Putting Capex on a Diet
December 8, 2008
The point to keep in mind is that Google has been working for a decade to build out its infrastructure. One of the benefits of the company’s willingness to tackle hard engineering problems is that Google obtains a better return on its hardware dollar. Data included in my 2005, The Google Legacy suggested that Google can spend a dollar and get as much as five times to performance that a non-Googlized data center would get. The data appeared in Google technical papers. Some of these papers were written by big Googlers; others by small Googlers. What the performance data share is information that provides a glimpse of the computing capability in Google’s data centers. If we flip the performance data around, a competitor would have to invest as much as five times what Google spends to get comparable performance. Is Google’s engineering that cost effective? Well, a five hundred percent performance boost may be optimistic, but when a data center can cost $600 million the implications are interesting. A competitor would have to spend more than Google to match Google’s performance on data manipulation, disk reads, and queries per second. Let’s assume that Google gets a 25 percent boost. For a competitor to match Google’s performance, the competitor would have to have the known bottlenecks under control and then spend another $125 million which makes a $600 million data center hit the books at $725 million. If you pick a larger performance boost such as two hundred percent, the $600 million data center will require $1.2 billion in capex to match Google’s capacity. Of course, no one would believe that Google wrings such a performance advantage from its commodity hardware. Competitors prefer branded equipment. What’s in the back of my mind is that Google may be keeping its cards close to its chest.
The Washington Post’s “Google Turns Down Some of NC’s Tax Incentives” explains that the economic downturn, among other factors, may be causing Google to trim its capital expenditures. The Washington Post here quotes a letter Google sent to North Carolina officials. For me the key phrase was:
While Google “remains pleased and committed to its Lenoir operations,” economic conditions make it too difficult to be sure the $600 million data center complex will expand as fast as previously thought, the letter said. “Yet the company fully expects to achieve employment and capital investment levels that are consistent with those that the state announced in 2007,” Charlotte attorney John N. Hunter wrote on behalf of Google.
The Google capex expenditures are going to become more important. The economic downturn is affecting most organizations, and I think the GOOG may be battening down its hatches. Good Morning Silicon Valley takes this position. You can read its take on the capex shift here.
What happens if Google does trim its capex for data centers? Maybe Microsoft’s new data centers will leap frog over Google? Google could find itself on the wrong side of high performance if Microsoft builds its own super performance innovations into its data centers. What the Washington Post makes clear is that Google is slowing down at least in North Carolina. The Google may be trying to trim costs by rethinking certain investments. This is another sign of Google’s increasing maturity and could indicate the opening that Microsoft needs to hobble the search Googzilla.
Stephen Arnold, December 6, 2008
Google: More Aggressive Sales in DC
December 4, 2008
The Washington Post here revealed that Google wants to move the US government toward Google’s application platform. The newspaper and Google called this cloud computing, but the notion is that the US government can save money and improve its work flow by letting Google handle the computers and the software. Government professionals can just use the GOOG. “Google Goes to Washington, Gearing Up to Put Its Stamp on Government” by Kim Hart does a good job of reporting a push that has been increasing over the last 12 months. For me the most interesting portions of the article are those which provide data about Google’s reach; for example, the District of Columbia has more than 30,000 people using Google’s services. Another comment that caught my attention was:
Google’s foray into government business is not only a sign of the company’s expansion into other industries, it’s also a sign of the changes underway in Washington’s technology landscape. New firms are moving in, branching out and making deals, perhaps beginning to blur the line between the robust government contracting world and the consumer-minded firms that continue to take chances and thrive.
My take on this is slightly different, which is normal goose behavior. Specifically, Google’s office is busy because government agencies are contacting Google to talk about the firm’s products and services. In my experience, Google needs only to answer email and the phone. The firm’s magnetic pull is responsible for the strong uptake of Google Search Appliances, Google Maps, and other services. Other companies have to work much harder to get in to see top executives. Google is a big deal and the Googlers are viewed as minor celebrities.
As the ad business growth slows, Googzilla will become more of a disruptive force in the government sector and in the enterprise. Companies that once ignored the GOOG will have to adjust and fast.
Stephen Arnold, December 4, 2008
Microsoft’s Cloud Stability
December 1, 2008
TechCrunch’s “Live Cashback Having a Bad Black Friday” is a gentle description of a cloud service outage at Microsoft. You can read the post here. The diplomatic write up said, “the [Microsoft] site has been down much of the morning [November 28, 2008].” The outage is a problem for Microsoft. I think it is an even larger problem for consultants who suggest that their expertise includes cloud computing. Microsoft is spending for what it views as the best technical talent in cloud computing. If Microsoft can’t pull off a service for its important Live Cashback system, are the self proclaimed Enterprise 2.0Â experts part of the solution or part of the problem? Unike the Enterprise 2.0 jargon, cloud computing refers to specific technical functions. Maybe Microsoft should turn over its cloud computing efforts to an Enterprise 2.0 wizard and fix the problem. Sound silly? Well, it’s not silly; it’s a dangerous and flawed idea. Engineers are needed, not MBAs with an inflated sense of their own abilities.
Stephen Arnold, December 1, 2008
More on IBM’s Cloud Play
November 24, 2008
Larry Dignan’s article “IBM Launches Cloud Computing Services” provides more color for what I perceived as a gray and wispy initiative. You can read his story for ZDNet here. He also provides a link to other ZDNet articles about IBM’s new initiative to regain territory ceded to Salesforce.com and other early entrants in the cloud computing space. For me, the most interesting comment in his article was:
Of IBM’s services, its security efforts may be the most interesting. IBM said it has cooked up a company wide project to create a security architecture for cloud computing environments. The project pulls together IBM’s research and security arms with the rest of the business. The idea is that enterprises provide and get cloud services with security at least as good as what they get in traditional computing environments.
When I read this passage, I thought of Oracle’s attempt to sell enterprise search based on the Oracle security functions. I don’t think lead balloons flew too well when SES10g lifted off, and I don’t think security will work much better for IBM. Here’s my reasoning, and you can provide your facts about why I am wrong in the comments section to this Web log:
- IBM has not been able to stick with a network services business. The company dabbled in Internet services and dumped the business to AT&T where it ran aground a decade ago. Need to refresh your memory? Click here. Neither IBM nor AT&T have proven themselves able to do much more than stick with well worn grooves. If you have ever been to Pompeii, you will see the grooves that carts made in stone streets. Once these companies hit a groove, change becomes really difficult. I know about the elephants dancing stuff, but I think cloud computing is not likely to be the winner for IBM that IBM hopes it will be.
- Success in the cloud requires capital investment. I don’t think that most companies appreciate the cost of the moving applications to “out there” on the Internet. Microsoft’s massive investments, if they are really made and supported, could sink the $65 billion in revenue giant into a sea of debt. Building plumbing is one thing. The real cost comes from upgrading and enhancing the infrastructure. IBM has not demonstrated that it can pull off this type of operational plan. The company has shifted its chip business because of the cost issue. I don’t think the company will have the appetite for infrastructure because it tried that meal once before and exited the business.
- The technology challenges for cloud computing are significant. Security is one modest component, but there are larger, far more troublesome thorns to pick out of IBM’s blue sweater. For example, the need to create multi tenant operations that deliver satisfactory performance and stability. Also, the challenge of engineering around the bottlenecks that old software like IBM DB2 throws into the face of licensees with far less demanding data tasks. In short, IBM has legions of engineers, but moving IBM inventions from the lab and making money from them is not an assembly line process. IBM is primarily a consulting firm, and its operational competencies continue to raise questions in my mind.
You may have greater confidence in Big Blue than I do. That’s fine. I don’t plan on shifting my company’s future to IBM’s cloud. In fact, I don’t think IBM realizes that its pal Google is going to be a big player in cloud services, which calls into question how Google and IBM will interact in this important new computing environment. My hunch is that the GOOG will dine on some IBM body parts. IBM’s elephant may be too big to notice that its tail was gobbled by Googzilla. An elephant can dance, but I sure hope it can fight too, not just trumpet.
Stephen Arnold, November 24, 2008
HP: Cloud Threat Looming
November 24, 2008
TradingMarkets.com published “HP Says It Powers 41 Percent Of Computers on Top 500 List.” You can read the full text of this article here. The title suggested to me that this was a routine PR type story, the French verb gonfler covers this type of pumping up action. For me, the most important comment in the article was:
Over the last several years, we’ve seen an explosive growth of blade servers for a widening range of high-performance computing applications – from digital media creation and online gaming to more traditional HPC applications such as computer-aided design,” said Earl Joseph, program VP, High-performance Computing, IDC Research. “Previously, customers’ only choice for HPC was a high-end, multi-million dollar supercomputer. Now, blades offer a highly flexible, scalable, lower-budget alternative to the proprietary systems that historically dominated the Top500 list.”
HP has reported good financial results and is one of a select number of companies to expect good numbers into 2009. But I looked at these figures and asked myself, “What happens if Ford, GM, and Chrysler go out of business or sharply curtail their investments. What happens if the Top 500 become the Top 400 or Top 300?” Then I asked myself, “What happens if the commodity cloud computing option makes significant inroads into the biggest companies who have to decide between their concerns about security and keeping the lights on?”
The news story suggests that today’s rosy tinted picture could turn gray. Branded hardware costs more than commodity hardware. But what really spikes the cost is the support and service that branded hardware drags like a snake moves its rattles through the sagebrush.
After reflecting upon this news story, I concluded that HP could find itself in a revenue pickle that cutting prices on its slick servers won’t be able to correct. Help me understand what I am missing?
Stephen Arnold, November 24, 2008
More about Microsoft’s Data Centers
November 23, 2008
Business Week has tilled an old field with its “Microsoft to Google: Get Off My Cloud” by Peter Burrows. You must read here Mr. Burrows’ description of Microsoft’s initiative to build out its data centers. The multi year effort under the firm hand of the national basketball association’s (if I understood Mr. Burrow’s undefined acronym) former technology wizard, Debra Chrapaty. Mr. Burrows picks up the story in 2003 when Microsoft started its initiative to build “20 supersize data centers that can cost as much as $1 billion apiece,” Mr. Burrows informed me that:
Rather than spend hundreds of hours opening server boxes, and connecting them with cables and loading them with software, Microsoft can roll in a container in just a couple of days. The hope is to run the facility with half as many people as at its previous sites. Even better, it’s easier to monitor and whisk away heat generated in these confined modules than cooling an entire building. One source says Microsoft hopes the design will help cut by one-third the power bills that typically take up some 40% of a site’s operating cost.
Mr. Burrows talks to a Googler who points out that Google’s been doggedly building infrastructure for a decade. The only issue I had is that Mr. Burrows did not seek out one of Google’s infrastructure experts, settlling for a marketing type in my opinion, not a Google research wizard like Jeff Dean of Simon Tong. I guess one Googler is as good as another for Business Week’s fact checkers. The Google source quoted in Mr. Burrows’ article understates what Google has in place. From Google’s point of view understatement is part of the game plan. For the reader, understatement makes Microsoft’s investments and chances look quite amazing.
The story ends with a contrast between Google’s and Microsoft’s approach to secrecy. Microsoft talks about what it is doing. Microsoft talks to Business Week, hence the story. Microsoft talks to Dell, which I am not sure makes Hewlett Packard too happy. Microsoft talks to developers, users, and even addled geese like me.
In my opinion, talk is cheap, a lot cheapter than buckling down and addressing the issues that Google has confronted and resolved to its satisfaction; for example:
- Google has addressed the need for speed in Web type applications. In fact, the GOOG can sort a petabyte of data in a few minutes, and that’s not easy to do even on Microsoft’s whizzy new super computers.
- Google has jumped over the database problems with its dataspace initiative, Chubby, and some innovations in tokenization.
- Google has long been a driver in commoditization and packaging of low cost computing horsepower and lights out data center operation.
I cover more of Google’s innovations from the period 1998 to 2006 in my Google studies. For our purposes in this opinion piece, look at the time span. Google’s been working for a decade, and now Microsoft is starting according to this article later in the game. As a result, Microsoft has to spend money to close the gap. Unfortunately, the closing of the gap does not include confronting the performance issues that will persist in Microsoft’s new data centers. These issues are related to the Microsoft architecture and server software itself.
One quick example. Two load balance, Microsoft has a clever design. The only hitch is that additional servers are needed to prevent bottlenecks at the points originally designed to reduce bottlenecks. The result is a need for additional servers. Does this information mean anything to Business Week? Nope.
It means something to Google, and it will mean something to Microsoft’s financial officers when the costs for resolving some fundamental architectural problems keep Microsoft’s data center investments in the category “hard to control.” If you have other information, please, use the comments section to put me on the true path to understanding. I have the Microsoft technical papers in front of me, and these are indeed quite interesting with regards to core engineering issues in the Microsoft data center initiative. Hardware, gentle reader, can’t solve some performance issues. That’s what makes Google a formidable competitor: software engineering, not commodity hardware.
Stephen Arnold, November 23, 2008
Google: Try This on Windows Server
November 23, 2008
Google has thrown down a gauntlet to the supercomputer crowd. You can read “Sorting One Petabyte with MapReduce” here. To learn more about MapReduce, click here. Now this recent Google gauntlet is digital, not one of those Sir Walter Scott fictional jobs with yellow tassels and brass fittings. Google is saying, “Take a terabyte of data and sort it. Now beat this time: 68 seconds. Once you have that whipped, take one petabyte of data and sort it. Beat this time: 362 minutes. You can read the details, get a comparison so you have a sense how much data a petabyte comprises, and even a touch of Googley humor. The sentence begins, “We are not aware…” If you don’t laugh out loud, well, you aren’t Googley. Why mention this type of lab rat exercise? Four reasons:
- In my opinion Google is reminding the folks who are yapping about how fast their supercomputers are that the GOOG is running a zippy computer too
- Make it clear to Microsoft that it has some work to do to match Google’s as is data center performance
- Show that Google can tackle big data as part of its real world applications. If you allow unlimited block uploads to Google Base, then you darn well have to whip that stuff around withoiut choking the services that pay the bills like ads.
- Put a benchmark in place so that competitors like IBM, Oracle, and Yahoo get a hint about how far ahead in data management Google is now. Today. This instant. IBM’s, Oracle’s, and even Yahoo’s technologists may be able to say their system is as good as Google’s. Google is too polite to come right out and say, “We’re faster across the board on certain key benchmarks.”
Oh, I am feeling frisky this Sunday morning. Here’s a thought for you: if you don’t care about these sort speeds, I guarantee that you will have a tough time understanding what the GOOG has been doing for the last decade while everyone talked about the company as an ad agency.
Stephen Arnold, November 23, 2008
Amazon, Capgemini: Cloud Computing Needs MBAs
November 21, 2008
SharePoint in the cloud is tricky even for Microsoft. Imagine that you are a Microsoft competitor working with a company stuffed full of MBAs and egos. Your job, if you decide to accept it, is to get SharePoint working in the cloud. One of my two or three readers sent me a link to a Web log on MSDN here. That blog post pointed to this Amazon news story which I had previously overlooked. If you want even more detail, there’s a PDF with more Amazon word smithing here. After you tunnel through the links, the point is that a consulting firm will sell and integrate client IT functions with Amazon Web Services. After all the clicking and backlink reading, Amazon’s stock fell even lower. I know the financial climate is bad, and I had hopes that this significant announcement about the AWS ecosystem would have revivified the Amazon share price. I was disappointed. Amazon is working hard to make its cloud initiative return big dividends. But the company counts objects, not dollars. Now the job of using Microsoft applications to compete with Amazon will have the able assistance of MBAs. I will watch this exercise with considerable interest.
Stephen Arnold, November 21, 2008
-
- Subscribe to Beyond Search
Feature archive
News archive
-
