Willy Everlern and Big Data Hype-ocracy

Posted on: September 4th, 2013 by Dan Graham 1 Comment

Willy Everlern is a young reporter at BigMedia.com who doesn’t understand data warehouses or computers.  His boss pushes Willy into many topics, so it’s hard for Willy to master any of them.  Even worse, Willy thinks compiling articles and one-liners from a bunch of internet blogs is research.  He doesn’t call the analyst firms, vendors, or customers for facts.  So some of Willy’s articles are speculations built on a firm foundation of hype.   Willy’s recent BigMedia.com article has gotten Mike-O, a PR JD (Jive Detector)  at Teradata, in a huff.

--Mike-O: “Dan, have you seen this Big Data: Beyond the Data Warehouse article at BigMedia.com?  It’s wrong on too many levels.  Looks like he’s diagonally parked in a parallel universe. Call this guy NOW.”

--Ring ring: “Hello, Willy?  This is Dan at Teradata.”

--Willy: “Oh, howdy Dan.  How’s everything going these days?”

--Dan: “Willy, we need to talk about your recent blog article.  There are a whole bunch of errors that have my customers calling us all confused. And you scared one investor to death.”

--Willy: “What errors?  I worked really hard on Big Data: Beyond the Data Warehouse. “

--Dan: “Well, let’s look at your article.  Skip to where it says “Database technology is ill-suited for big data.”   That may be true of most databases, but it’s not true of Teradata databases.  Our core competency has always been scalability to the largest databases in the world for over 30 years.  Our Petabyte Club of customers has over 50 member installations.  One of those machines is over 60 petabytes in size.  And our Aster database is chugging along with some 800 terabyte installations.  That’s big in most people’s thinking.  I can get you reference calls with some of these customers.  You should also check out Ventana’s blog on Teradata Addresses the Foundation of Big Data Analytics.1

--Willy: “Yeh, but it’s not the new Map Reduce stuff.  You missed my point.”

--Dan:   Well if Map-Reduce is your definition of big data, Gartner says Teradata Aster is one of two databases that implement MapReduce directly inside a DBMS. 2  Actually Aster is the first database with full map reduce in it. (Psst Willy – keep this a secret but we share technology between Teradata Database and Aster. Think about it.)  So we’ve got scale-out AND Map-Reduce in our databases. I admit, only one other vendor can do that so you weren’t completely off base.  But Teradata is an exception when it comes to big data. It’s our specialty.”

--Willy:  “OK, I’ll give you this one.  You got me.”

--Dan: “Just a little advice Willy – never buy sushi from a vending machine.”

--Willy: “What?”

--Dan: “Forget it.  I’m just playing with your head.”

--Dan: “So Willy, there’s another thing to discuss.  Skip to where the article says “Databases can’t handle multi-structured data. Willy, the majority of databases can’t handle tweets, web-surfing logs and internet-connected sensors.  You got that part right.  But Teradata and Aster databases are an exception again.  They peel apart multi-structured data easily.  Remember that 60 petabyte Teradata system I mentioned?  The customer uses it to unravel weblogs into name-value pairs and then do analysis of consumer purchasing behavior. Weblogs are as unstructured as it gets – the stuff looks like a cosmic hairball of data.  Aster’s SQL-MapReduce goes even further.  Aster can actually do joins of Twitter data, Facebook data, and consumer history to correlate patterns across them.  In English that means Aster can look at social data and tell you when a consumer is gonna jump to the competition.”

--Willy: “Wow.  So your aging old databases are really some kind of modern social network movie and The Matrix all in one!”

--Dan: “Slow down Willy – you’re scaring me.  Let’s stick to the facts.  I’ve got one more topic and I’ll let you go.  Now look in your blog where it says “Data warehouses are inflexible, resistant to change.”  Willy, I’m going to agree with you on this one, but it’s not what you think.  It’s not a technical problem, it’s a people problem.  It affects all the database vendors.  Some DBAs have become data jailors, keeping the data locked up.  And some BI governance committees have gone too far.  Business users could get more of what they want if they adopt Agile development methodologies. Plus Teradata built something called Data Lab that’s really flexible for A/B testing and new ‘what if’ ideas.  Companies can prototype something in a few weeks and promote it into production in another couple weeks. It helps, but we still need people to adopt the Agile methodology3.  There’s nothing wrong with the database software.  We could use your help getting the message out on Agile.”

--Willy:  “Hmmm.  Well, I wish I’d known all that a few weeks ago. How was I supposed to know?”

--Dan:  “Willy it’s my job to run and fetch whatever you need.  Anytime you want to mention Teradata, just call and I’ll be working for you.  Seriously, Mike-O pumps Barry Manilow music into the office for 3 hours anytime he  spots “jive’’  from one of you bloggerazzi  guys.”

--Willy: “Oh my god – that’s torture.  Look, I’ll call you next time if you can point me to the analyst stuff that makes me look smart.  Thanks Dan.”

--Dan:  “I’ll be glad to help, Willy.  Three hours of ‘Oh Mandy’ is highly motivational stuff.”
Later.

--Mike-O, PR JD: “Well, will he ever learn?”

--Dan: “I don’t know.  I don’t know.”

 

1 - Ventana, Tony Consentino, http://tonycosentino.ventanaresearch.com/2013/05/03/teradata-addresses-the-foundation-of-big-data-analytics

2 - Gartner, IT Market Clock for Database Management Systems, 2012, September 2012

3 - TDWI, Benefits of Agile Data Warehousing: A Real-World Story, July 2013

One Response

  1. Martin Willcox

    September 5, 2013

    Great work, Dan! Hopefully the next PB scale benchmark we run for a customer will cut the power and spare you from Barry M!

    Reply

Leave a comment

*