New job for mainframes: Hosting private clouds

There isn't as much excitement using tried and true technologies.  In conversations with your peers, they tout using the "latest and greatest" technologies.  Then you ask them how they are running.  How is their security?  How is performance?  They aren't as excited when they answer.  Systems go down, data is hacked, response times vary...

Article after article talks about and company after company is looking more to mainframes for their power, performance, security, and up-time.

In a recent article in Computerworld, the utilization of the mainframe as a cloud was once again identified as a great solution.  Quoting Tam Harbart:

"Nevertheless, mainframe vendors contend that many companies want to use their big iron for cloud computing. In a CA Technologies-sponsored survey of 200 U.S. mainframe executives last fall, 73% of the respondents said that their mainframes were a part of their future cloud plans."

Judith Hurwitz,president and CEO of Hurwitz & Associates expands on the strength of the mainframe as a private cloud:

"You have this incredibly scalable server that's very strong in transaction management.  Here's this platform that has scalability and partitioning built in at its core.  Plus, the mainframe's strongest assets -- reliability, availability, manageability and security -- are the very characteristics that companies are most concerned about as they consider rolling out major business applications in the cloud." 

The excitement in these solutions is they work and they are secure and, at the end of the day, aren't those two of the most important things you can possibly ask for?

 

Business Intelligence is Critical - Relational is better!

In a recent article by gigacom, author Derrick Harris points out that Amazon has invested significantly in analytic database startup ParAccel, whose CEO, Chuck Berger, told Harris that “Amazon sees significant value in what ParAccel does.”

In addition, Gartner recently surveyed CFO’s about IT investments.  Business intelligence (BI) is the top technology initiative from the perspective of the senior financial executive.  For a combined 65 percent of choices, BI ranked as the technology with the highest demand, while 46 percent ranked enterprise business applications, such as enterprise resource planning (ERP) and integrated financial management solutions, as investment priorities.  When viewed within the larger scope of operations' infrastructure, however, business applications (30 percent) were seen as more important than BI (23 percent) in 2011.

ParAccel  makes a columnar (as opposed to the traditional row-based model for relational databases) database designed for fast analysis of large amounts of information.

This reflects a complete misunderstanding of database theory and implementation.  Relational theory is rock solid.  It is calculus, it is math.  Columnar can be an implementation of relational - sort of like the specific formula for a specific cross section of a specific wing of an airplane.

One could implement a relational database with a columnar storage and still be relational.  Or, one could ignore relational and do a non-relational columnar storage.  Of course, since relational is superior, most columnar implementations preserve the calculus of relational theory - and use SQL.

The funny thing is a person with decent knowledge of a relational database can implement columnar storage, say with Oracle.  But most people haven’t figured this out.

The advantage, perhaps, to the columnar only databases is they are optimized for columnar queries. You can still get close in performance if you know how to "tune", but for people who do not know how to tune, columnar forces them into more optimal structures. The best analogy is the race track ride at disneyland that forces pre-teens to stay in their lane.  Modern relational databases are like off road vehicles.

So, even though in some cases columnar is better, it is unacceptable for the majority of database tasks.

Why make a mad dash to reinvent the wheel?

PC World published an article on June 23, 2011, stating four companies are going to rethink their databases for the cloud.  NimbusDB, Xeround, ParAccel, and Cloudant are all developing new database technologies "to solve what they see as the shortcomings of traditional, relational database management systems in a cloud environment. "

If you read the article, they are all circling around reinventing something that will cost at least $100M and, unless they spend hundreds of millions and build their own mainframes it still leaves them stuck on Windows or Linux.  How secure is that? Where will they really be at the end of that?

It is true they are the shiny new car, but how well do they run?  How secure is your data?  Hackers are getting smarter and smarter and they have found the flaws in Windows and Linux.  Why not run on a database that executes in read-only so your data can't be hacked.  What price peace of mind?

Old Tech Loses Its Glow

Why they need Dark Star  Cloud....

Old Tech Loses Its Glow

Microsoft, Cisco Aren't Trading Like Growth Stocks; Valuations Fall Below S&P 500

By JONATHAN CHENG, MATT PHILLIPS and STEVEN RUSSOLILLO

In their lust for the latest technology IPO, investors appear to have ditched some of their former tech darlings.
Amid the heavily hyped chatter on initial public offerings from Zynga, LinkedIn and Facebook, seemingly old-school tech companies such as Apple and Google were conspicuous laggards during the second quarter. The Nasdaq Composite index ended the quarter in negative territory, compared with gains for the rest of the market.

The buzz over the latest crop of tech newcomers has been palpable. LinkedIn ended its first trading day with a stock-market value of about $9 billion. On Friday, Zynga filed paperwork with regulators for a $1 billion IPO, although the final amount could be considerably different. Facebook is aiming for a stock-market value of more than $100 billion, placing ahead of the likes of Cisco Systems and Hewlett-Packard.
Still, some investors say the pummeling of tech stocks may have gotten out of hand.
Investors in LinkedIn are paying about $21 for each dollar of the social-networking website's estimated 2011 revenue. Yet they are paying less than $6 for each $1 of expected revenue at Google and Apple.
"The reality is Apple and Google are large-cap growth stocks, but they're not trading like such at the moment," said Paul Bard, director of research at Renaissance Capital, an IPO investment-advisory firm. "Investors aren't treating them as growth stocks. Instead, they're placing their bets in this new wave of Internet companies."
By one measure, tech stocks are at their cheapest compared with the rest of the market in almost 20 years.
For the past four months, the price-to-earnings ratio of the tech sector has been below that of the S&P 500, according to estimates compiled by UBS. That's the first time that has happened since 1992.
As of last week, the S&P 500 was trading around 12.4 times one-year forward earnings, and tech was fetching 12.1 times, UBS estimates. Apple stock is now fetching about 12 times earnings. Google is trading at about 13.8 times. Intel and Microsoft have P/Es of less than 10.
The turning point for tech came in March, as Japan's earthquake and tsunami caused supply hitches and worries emerged about global economic growth, dragging tech stocks down more than most.
Jonathan Golub, chief U.S. equity strategist at UBS, thinks now is the time to jump into tech stocks, pointing to a number of tech names he calls "unusually cheap."
Some say valuations have gotten so low that they are likely to rise should there be any stronger signs that the economy isn't mired in a prolonged slump. A few glimmers of hope for those bulls came last week, and tech recovered along with the rest of the market.
Even if the economy does sputter, the valuable—but underappreciated—combination of cash flow, accumulated cash and decent dividends offered by companies such as Microsoft, Cisco, Intel and Hewlett-Packard could become attractive to conservative investors.
Mark Luschini, chief investment strategist at Janney Montgomery Scott, suggests there may be some early signs those attributes are gaining notice. He points out that Microsoft shares rose 2.4% during the second quarter, handily outpacing the S&P 500's 0.4% decline.
"They've been sort of unsponsored, if you will, almost orphaned in the marketplace," Mr. Luschini said of large-cap tech stocks. "I think people are looking at these things as if they're some sort of old-world victim of some new-age technology that make these franchises no longer viable. But I think it's just the opposite."
Not everyone is so sanguine on the sector. Burt White, chief investment officer at LPL Financial in Boston, notes that the tech sector's earnings and revenue growth have lagged that of the broader market in recent quarters.
"That's not what we expect from tech," Mr. White said. "The market is just concerned that it's not going to get the growth rate from tech that we thought we would."
Investors are regularly willing to pay top dollar for earnings if they think a company has significant growth potential. But with that seemingly not the case for former high-fliers like Cisco and Hewlett-Packard, their shares are among the worst performers this year among the 30 Dow Jones Industrial Average components.
Cisco has slid 22% through Friday, H-P has dropped 12% and Microsoft has lost 6.8%. U.S. markets were closed Monday for the Independence Day holiday.
"If growth doesn't materialize, these valuation multiples can come crashing down," Mr. Bard says. "But if growth materializes, history has told us that somewhere down the road the multiples will normalize to levels of more mature growth companies."
Phil Orlando, equity strategist at Federated Investors, says investors now have a rare chance to buy some formerly expensive stocks at reasonable valuations.
"The cheapness is an opportunity to buy," Mr. Orlando says.
Write to Jonathan Cheng at jonathan.cheng@wsj.com, Matt Phillips at matt.phillips@wsj.comsteven.russolillo@dowjones.com and Steven Russolillo at