Sun’s New X4450 Servers

September 1, 2008

I have a soft spot for Sun Microsystems hardware. Heck, we even like Solaris. We love mounting devices and performing the Sun certification process before deploying hot new hardware. I learned that Sun Microsystems released a “white paper” about its newest servers. If you are a hardware junky as I am, you will want to click here and wallow in the technical goodness of these gizmos. The white paper is “Sun Fire X4150, X4250, and X4450 Server Architecture.” What I found interesting was this statement from page 38:

Organizations strive to reduce variety of platforms in the data center, even when a wide range of workloads are present. To help this effort, the Sun Fire X4150, X4250, and X4450 servers can run the Microsoft Windows operating environment. Indeed, these servers have passed stringent Microsoft compatibility test suites, achieving ā€œDesigned for Windowsā€ certification. This certification demonstrates Sunā€™s commitment to providing the best platforms to run not only the Solaris OS and Linux, but Windows as well.

I would love to have a couple of XX4450s. Finally a hardware platform that can make Vista hum. Now about those SQLServer back up time outs. I am not sure even a X4450 can swizzle away that issue. I am confident that I could eliminate some of the latencies we have encountered with certain blue chip search systems’ content processing sub systems with a rack of these X4450s stuffed with memory and upscale storage gizmos. The ceilings of 50 million documents would remain, but I would hit that ceiling more quickly I surmise.

Stephen Arnold, September 1, 2008

Enterprise Social Software: Security in a Single Word

September 1, 2008

Disclaimer: I write for KMWorld.

My feedreader delivered a story from KMWorld with the intriguing title “Enterprise Social Software Technology.” You will want to read the article here. It provides a glimpse of what I call Heidi Klum (the star of Project Runway and international supermodel) “one day you’re in and the next day you’re out” approach to enterprise software. Social software is the buzzword for software mechanisms that streamline certain types of communication and make it possible to assemble information from different sources in a single Web page. The write up does a good job of collecting buzzwords and explaining each within the context of an enterprise. You can read the full text here. What struck me as interesting is that the issue of security gets a single word. With Sarbanes Oxley, BASEL II, and other governmental regulations increasing the grip, the importance of information security ratchets upwards a notch. Organizations quick to embrace Webby solutions may find themselves scrambling to make whizzy systems mesh with regulatory and legal guidelines. When a legal matter confronts an organization, the importance of “security” pops up a level. Communications germane to a legal matter may be “discovered”. One hopes that “security” warrants more than one word, which is what the author and KMWorld allocate to this topic. When your organization deploys social software, how will the messages be archived and made findable. What’s the audit trail for spoofed messages? What actions are needed to ensure that confidential information finds its way into the shared information space? Does your organization’s security methodology have the means to deal with content germane to a hot topic such as a clinical trial result, an employee’s health, or a new product that is critical to the company’s revenues? Non-social content management in most organizations is a disaster. Will social software tidy the messiness or accelerate entropy? At my age, I err on the side of planning, caution, and careful consideration of regulatory and competitive issues. I leave the craziness to parvenus, wily MBAs, and those with more enthusiasm than common sense. Agree? Disagree?

Stephen Arnold, September 1, 2008

Google Ecosystem: An Appliance Appliance

September 1, 2008

Network Box has released a web content filtering technology to support Google Safe Search and Safe Browsing functions. Device Management Forum’s “Network Box Releases Web Content Filtering System for Google Safe Search and Safe Browsing.” You can read the story dated August 25, 2008, here. According to the story,

Network Box USA customers will now have a content filtering engine that uses Google’s Safe Browsing system to check URL requests against Google’s constantly updated blacklist of suspected phishing- and malware-infested web pages. These web filter categories will be in addition to Network Box’s existing anti-virus and anti-phishing solution, which uses Kaspersky and ClamAV to inspect pages for malicious content.

With a bit of work, this “box” could be used in tandem with Google Search Appliances and other functionality added to bolster the GSA in certain environments. An appliance for an appliance strikes me as an interesting innovation in the Google ecosystem. You can learn more about Network Box. Mind the url, please. http://www.network-box.com/

Stephen Arnold, September 1, 2008

Citation Metrics: Another Sign the US Is Lagging in Scholarship

August 31, 2008

Update: August 31, 2008. Mary Ellen Bates provides more color on the “basic cable” problem for professional informatoin. Worth reading here. Econtent does an excellent job on these topics, by the way.

Original Post

A happy quack to the reader who called my attention to Information World Review’s “Numbers Game Hots Up.” This essay appeared in February 2008 and I overlooked it. For some reason, I am plagued by writers who use the word “hots” in their titles. I am certain Tracey Caldwell is a wonderful person and kind to animals. She does a reasonable job of identifying problems in citation analysis. Dr. Gene Garfield, the father of this technique, would be pleased to know that Mr. Caldwell finds his techniques interesting. The point of the long essay which you can read here is that some publishers’ flawed collections yields incorrect citation counts. For me, the most interesting point in the write up was this statement:

The increasing complexity of the metrics landscape should have at least one beneficial effect: making people think twice before bandying about misleading indicators. More importantly, it will hasten the development of better, more open metrics based on more criteria, with the ultimate effect of improving the rate of scientific advancement.

Unfortunately, traditional publishers are not likely to do much that is different from what the firms have been doing since commercial databases became available. The reason is money. Publishers long to make enough money from electronic services to enjoy the profit margins of the pre digital era. But digital information has a different cost basis from the 19th century publishing model. The result is reduced coverage and a reluctance to move too quickly to embrace content produced outside of the 19th century model.

Services that use other methods to determine link metrics exist in another world. If you analyze traditional commercial information, the Web dimension is either represented modestly or ignored. Mr. Caldwell’s analysis looks at the mountain tops, but it does not explore the valleys. In those crevices is another story; namely, researchers who rely on commercial databases are likely to find themselves lagging behind those researchers in countries where commercial databases are simply too expensive for most researchers to use. A researcher who relies on a US or European commercial database is likely to get only an incomplete picture.

Stephen Arnold, August 31, 2008

Google: A Great Place to Work

August 31, 2008

If you want to refresh your memory about how wonderful Google is to employees, you will want to read the Red Orbit “Way of Life in the Google Complex” here. After a summer of transparency, the writer–possibly a Googler or a PR maven in disguise–reprises the wonders of Google. You get a reference to the lava lamp. You get a reminder about the grand piano. You get it all. The writer leaves out entertainment like Tony Bennett at lunch, but you can revel in remarks like this one:

Most of the walls and dividers are made of glass so that rather than becoming a labyrinth of cubicles the buildings remain open and light is easily filtered through.

Yes, metaphorical transparency. That’s a nice rhetorical touch. Plus, I think it’s super that the author knows that the GOOG spends $72 million a year on these and other amenities.

Now that that the summer of transparency is nearing its end, Google’s fall campaign seems to be back to its wild and crazy math club ethos.

What a relief for me. I was growing tired of technical explanations, Google management’s advice to other companies about innovation, and talks that run the Google game plan. I hope that lovable Googler Cyrus somebody who told me and then others that a Google patent application drawing in one of my lectures was a Photoshop fake keeps retelling that fib. The lousy patent illustration was crafted by a Google wizard, not me. But Googlers don’t know what their own employer puts in its patent documents. Who wants reality to intrude on Google’s presentation of its world.

Reality, when viewed through lava lamps, is often different from “regular” reality, at least for me. Google’s lawsuits, Gmail outages, and plans for outer space made the summer of 2008 interesting to me as I watched this most important company enter its 11th year in business. Red Orbit’s write up is a useful glimpse into the world that Google wants me to believe exists. Do Googlers sleep on those fluffy animals instead of going home? Let me know if you have some insights.

Stephen Arnold, August 31, 2008

Google Maps Attract Flak

August 31, 2008

Google inked a deal with GeoEye to deliver 0.5 meter resolution imagery. One useful write up appears in Softpedia here. The imagery is not yet available but will be when the GeoEye-1 satellite begins streaming data. The US government limits commercial imagery resolution. Th Post Chronicle here makes this comment, illustrating the keen insight of traditional media:

Google did not have any direct or indirect financial interest in the satellite or in GeoEye, nor did it pay to have its logo emblazoned on the rocket. [emphasis added]

In my opinion, Google will fiddle the resolution to comply. Because GeoEye-1 was financed in part by a US government agency, my hunch is that Google will continue to provide geographic services to the Federal government and its commercial and Web users. The US government may get the higher resolution imagery. The degraded resolution will be for the hoi polloi.

Almost coincident with news of this lash up, Microsoft’s UK MSN ran “UK Map Boss Says Google Wrecking Our Heritage.” You can read this story here. The lead paragraph to this story sums up the MSN view:

A very British row appears to be brewing after the president of the British Cartographic Society took aim at the likes of Google Maps and accused online mapping services of ignoring valuable cultural heritage. Mary Spence attacked Google, Multimap and others for not including landmarks like stately homes and churches.

The new GeoEye imagery will include “valuable cultural heritage” as well as cows in the commons and hovels in Herfortshire.

Based on my limited knowledge of British security activities, I would wager a curry that Google’s GeoEye maps will be of some use to various police and intelligence groups working for Queen and country. Microsoft imagery in comparison will be a bit low resolution I surmise. MSN UK will keep me up to date on this issue I hope.

Stephen Arnold, August 31, 2008

No Google Killer Yet

August 31, 2008

I think it is still August 30, 2008, here in the hollow. My newsreader delivered to me a September 1, 2008, article. DMNews is getting a jump on publishing in order to make a picnic. The authors are a team–Ellen Keohane and Mary Elizabeth Hurn. You can read the article here.

The main point of the article is that Google is the leader in search. There were two interesting points for me.

First, the authors identified a search engine of which I knew not–UBExact. The url is http://www.ubexact.com. I ran one test query. I selected the geographic search for Louisville and entered the term “blacktop resurfacing”. The system generated zero results. I will check it out in a couple of months.

Second, the duo made a comment I found intriguing:

And, as with Wikia Search, Mahalo,OrganizedWisdom.com and Scour.com, UBExact also uses humans to improve the search experience. Human editors are contracted to eliminate spam, malicious content, unwanted ads and dead links and pages, Stephenson said. In addition to vetting content, the conĀ­tractors also organize Web sites based on content so users can search on UBExact by category.

Humans are expensive, and it will be interesting to see if privacy and click fraud impair Google. Oracle SES10g pitched security. Customers did not value security, and I’m not sure if UBExact’s hooks will either. Agree? Disagree? Let me know.

Stephen Arnold, September 1, 2008

Business Intelligence * Hots * Up

August 30, 2008

I was not going to read this VNU article. The phrase “hots up” annoyed me. One of the engineers in my rural Kentucky redoubt told me to take a look. His thought was that VNU was running a content free news story. I did. The story is by Rosalie Marshall, who is probably a warm and caring individual. The article is not completely content free. There’s a sales pitch tucked inside the sentences. And the title reveals a keen sensitivity to language; to wit, “Business Intelligence Hots Up.” I wish I could turn a phrase like that. The main idea of the article is a synopsis of findings by a research firm. The idea is that business intelligence is what customers desire. There’s a reference to Endeca, a company that has been trying to get more traction in business intelligence for several years. IBM gets a mention. Even Google warrants a comment too. For me the most important point in the article is the notion that business intelligence is becoming important. My thought is that search has not delivered. Vendors now chase revenues with business intelligence pitches. If you want to read this “hots up” stuff, click here.

Stephen Arnold, August 30, 2008

Microsoft: Another Search Buy

August 30, 2008

Microsoft has gathered another search system for its information retrieval basket. There’s a good summary on Yahoo News from Thomson Reuters. I will provide thisĀ  link, but it will go deadĀ  in a short span of time. Click here to see if you can access “Microsoft Buys Ciao.com to Boost E-Shopping Search” by Georgina Prodhan and a carton of contributors. The idea is for Microsoft to do a better job with shopping search. Microsoft lags Google, and the hope is to narrow the gap between Microsoft and the GOOG. For me the most important point is the article is:

Microsoft’s Mangelaars acknowledged the distance Microsoft had to cover, especially given the commercial edifice rapidly being built by online advertisers whose models depend on Google’s particular view of the Web. “It’s a race,” he said, “but we also believe it’s very early days in search technology.”

In my opinion, I am tired of hearing that it is early days for search. Search has been around since the 1960s. Sure, I’m on record saying, “Search sucks.” But whether search sucks or not is irrelevant when one company has a 70 percent share and a competitor has been trying to catch up for a decade. Leap frog, not me too, is needed.

Stephen Arnold, August 30, 2008

Enterprise Search Storage Estimator

August 30, 2008

Solrhack has posted a handy “rule of thumb” estimator for capacity planning. You can read the article and see the formula here. The article is called “Enterprise Search Capacity Planning.” Keep in mind that the multiplier can vary. Some vendors with excellent compression methods can generate indexes that are one quarter to one half the size of the total corpus processed. If you know of other tools like this, please, use the comments section of this Web log to share them. I will add them to the ArnoldIT.com and the New Idea Engineering page of search tools. Oh, a happy quack to Solrhack as well.

Stephen Arnold, August 30, 2008

« Previous PageNext Page »

  • Archives

  • Recent Posts

  • Meta