Adam Gartenberg's Blog

Business Analytics and Optimization, IBM and Social Marketing

Live Blogging Data Management keynote


I'm sitting in the Data Management keynote at IOD EMEA with Arvind Krishna, and will do my best to live blog it.  (And I figured out how to get on the wireless network now, so it actually will be "live."  Hit refresh for updates).

When look at history of data, has been around for 40+ years.  First db system literally built to help put man on the moon.  IMS - 48th anniversary - was built to ensure lunar landing went successfully and help people manage all the elements in supply chain and give astronauts answers they need immediately, as time delays could have catastrophic impact.

Then come to relational databases in 1970's and on to XML today.

Pressures for real-time data in logistics, in finance, everywhere are continuing.

At end of January announced acquisition of SolidDB.  In-memory database.  Latencies measured in microseconds.  Very high availability and very low cost.  Over 3M deployments worldwide.

Also investing heavily organically.

When look across all of these investments, it's all about how we return this investment to the values you seek to run your business well.

When we look at the pressures you face, what are we hearing back?  1.  How do we get more efficient about how  we use the data, lower the cost of deployment, get more agile and smarter in sense of more productive - do it in less time.  On the other side, everyone wants to manage data through lifecycle.  From when created to how deploy it, how performance tune it, and how archive it.

Have to manage data through lifecycle, not just the data that's inside.

IDS - Informix Dynamic Server 11.5.  Innovative use of cluster technologies to make sure system stays up.  It also often requires very few people to manage and run it.

DB2 9.  Databases are great at storing data.  The relational model is great, but it implies that you take the data, you form it up into columns, etc.  It's not easy to take XML documents, which is becoming the primary way we're working today.  XML doesn't naturally translate to this.  But if you could take something, do no development, and put it in database and run all semantics of databases and run intelligence against it, that's a value that's good for you.  ADP Nederland BV is leveraging that value of that.  Lower development cost, improves agility, don't have to shred data - can keep it as is, with digital signatures, etc.

InfoSphere Warehouse 9.5 - Heard from Fiducia this morning.  If look at how people use warehouses for sake of reports, have a number of problems.  1.  data is out of date almost as soon as form it.  If take data and put it somewhere for warehouse, will be a week late, a day late, but always late.  Create of datamarts, each one for own report.  Desire would be to have one warehouse, run all analytics on it without copying data out to data marts.  Reduce complexity of managing data.  Also need workload management.  Pleased to announce DB2 9.5.1 and InfoSphere Balanced Warehouse C4000.

IMS 10 - Accelerate integration into SOA.  Allow people to leverage investments made in systems over last few decades.

DB2 Optimized for SAP - Innovation in how we work with partners.  It's about partnership, integration, innovation.  It's not about partnership for our sake, it's for our customers' sake.  Roadmap of integration so is seamless when they come together.  We have joint teams in Waldorf and Toronto.  
Siemens video - Moved to DB2 to get improved performance.  Before had to use dedicated server, and moving to new server was big cost.  Now can use several partitioned server, can easily add new servers. Reduced total cost of ownership of SAP; expect to reach savings of 25%.  Saw 1-year ROI.  Integration was finished on time and on budget.

{SAP executive on stage}

Unlocking Business Value
Start on bottom with data mgmt and content mgmt.  Build on that with info integration, warehousing and management.  Can go on from there for business intelligence and performance management.  Closed acquisition of Cognos earlier this year.
Video from Romulus - BP - Provides business intelligence solution for restaurants, bars and pubs.  Helps them stock inventory and schedule staff based on expected attendance and drinking patterns - e.g., how does demand change if a football game is going well vs. not going well.  
On stage - Benefits from Cognos and IBM coming together.  Took one process from 18 hours to 1.5 hours.  Makes price point much better, but as a BP also see new revenue stream that weren't expecting... historically only sold Cognos SW, and now are selling IBM, too.  

Integrated data management - When look at overall task of deploying data.  As an industry we in the past have not been very good at providing an integrated environment to manage this lifecycle.  Think of the analogy of an integrated development environment that programmers have used for decades now.  The equivalent of that for the data world, even though data occupies major part of infrastructure, has not been there.  Our goal is to provide single environment to manage all data across all repositories and across entire lifecycle.  

How do we create, test, make subsets of databases?  How do we configure and do change management, performance tuning and triage.  And it's not just a vision - it's quite real.  {Graphic image of lifecycle model - I'll work to get it posted later today}
IBM Data Studio - One environment for developers to use their favorite environment to connect to the database.  Slash dev time up to 50%, simplify and speed development by finding errors up front rather than waiting until test it.  And optimize all of those queries through technology called pureQuery.  Seeing up to 20% improvement in java and db performance, and done in a way that's transparent to java developers and dbas.  Publish any web query so can be used in SOA environment.

Optim - How dig up old records? How control what records can be removed?  Optim help with this.  When doing testing, why need to make copy of all records vs. pulling out representative data. Yorkshire Water reduced batch processes by 75% (from 250 hours to 65 hours) by archiving historical data.

Web 2.0 - Social collaboration that we see.  Facebook, MySpace, YouTube.  What concepts can we bring from there to make our lives easier with the things we're trying to do?  Bring mashups to this world.  Can quickly see across 18 databases what is going on.

{End of session]