• Series 2 - Behind the Bar with the DB2 Cocktail Hour
• Series 3 - On the Bar Stool with the DB2 Cocktail Hour
• Series 1 - The DB2 Cocktail Hour
• Disclaimer and Copyright
• The podcast technology

11/15/2006 - Series 1 Podcast 18 Predictive Analysis (back to series 1), (back to top)

notes by Dan...Once again we are in our home away from home, Baltimore. We are drinking Korbell Champagne courtesy of our special guest, Annie Demosthenous! Today, in the last podcast of the series, we are talking about predictive analysis. These are highly effective techinques of measuring and reporting on database performance for a particular application, before the application is written. This includes tools provided by IBM, but also additional work that you can do yourself to get an idea about performance. Quit guessing and let the database tell you how to design the database!!! The song beds include Stranded and Disarray by Mohair, and Calling all Angels by Train. Dig it!!!

Disclaimer and Copyright

In this podcast we talk about:

1. Explain
2. Visual Explain
3. DB2 Estimator
4. Simulating Access, Generating Statements and Data
5. Program Tests (and some results)

Right click on the file name, and do a save as. Then you can listen on your PC, or download into your portable device to listen to anywhere! db2_Cocktail_Hour_118_preda.mp3

10/03/2006 - Series 1 Podcast 17 Live from IDUG; Trends and Directions in DB2 for z/OS (back to series 1), (back to top)

notes by Dan...Here we are from IDUG Europe, in the presidential suite of the Hilton! A wonderful group of people, and a bavarian theme including the singing of Ein Prosit! We are joined by David Simpson of TransUnion and Kurt Struyf of Competence Partners in a discussion about what's happening with the DB2 for z/OS platform. We discuss the move to version 8 of DB2, the direction of mainframe usage that we see in the industry, Java usage on the mainframe and elsewhere, and zIIPs and zAAPs! The song beds include Geile Zeit and Perfekte Welle by Juli. Grab a beer for this one!

Disclaimer and Copyright

Right click on the file name, and do a save as. Then you can listen on your PC, or download into your portable device to listen to anywhere! db2_Cocktail_Hour_117_tad.mp3

09/13/2006 - Series 1 Podcast 16 Top 5 SQL Performance Features in V8 (back to series 1), (back to top)

notes by Dan...From our home away from home in Baltimore! Many DB2 for z/OS shops are finally making the move to version 8. Some may be wondering what can they take advantage of immediately, and how to use it the right way. We have taken our now two and one half years of V8 experience, and have come up with our top 5 SQL performance features. Susan and I are drinking Vueve Cliquet Vintage 1999 Rose (nice), and her mother (yes, that's right) is not drinking but reads of the top 5 for us David Letterman style! A final note; I mentioned that I was testing multi-row fetch in an online environment, and was getting something like an 80% CPU savings. The actual savings after further testing was about 25%. The song beds include Would I Lie to You by the Eurythmics, Careful With That Axe Eugene by Pink Floyd, and Get What You Give by Tesla. During the podcast I made reference to a couple of recursive SQL statements. Here they are: generate the number 1 to 10, info rollup money, info

Disclaimer and Copyright

Top 5 V8 SQL Performance Features:

1. Recursion
2. Common Table Expressions
3. Multi-Row Fetch
4. Scalar Fullselect
5. SELECT from INSERT

07/13/2006 - Series 1 Podcast 15 Top 10 Ways to Waste CPU (back to series 1), (back to top)

notes by Dan...Ah, the pause that refreshes...We took a little break from doing podcasts after the big IDUG show, but there was no break from working! In fact we have been to more DB2 installations, and have performed more performance audits in the last two months then ever before in the same time period. This inspired us to create a top ten list of things to avoid doing if you want to save CPU cycles in your DB2 for z/OS environment. This podcast finds us in beautiful Conshohoken, Pennsylvania (can you gues who we're working for?). We're both drinking vodak on the rocks, and the music is all Sweet! That's right, the classics 70's rock legends! We open the podcast with, of course, Ballroom Blitz, the bed for the commercial is Action, and we close out with Fox on the Run. Oh yeah, baby!

Disclaimer and Copyright

Top 10 ways to waste CPU:

1. black box I/O generic design
2. too many indexes
3. random distribution
4. blind updates
5. RELEASE COMMIT
6. inefficient predicates
7. overuse of DISTINCT
8. processing of data in a SQL statement
9. excessive locking
10. overuse of compression

05/08/2006 - Series 1 Podcast 14 Live from IDUG; Legacy Migrations (back to series 1), (back to top)

notes by Dan....We had a dream (of course it came about one night after a few beers). Let's throw a party at IDUG and do a live podcast! We did it!!! A room full of people, a full bar, a ton of food (thanks to our dear friends that supplied the food), a sound system, and the live cocktail hour is born! Thank you to everyone who came to hang out, a special thank you to Lisa D for helping us set up, and to Chris Skelnik for helping with the sound. In this podcast we discuss legacy migrations. It's one of the biggest things this year on the z/OS platform. We've been there and done that several times, and share the secrets to success. We played Jetsetter and Nth Degree by Morningwood, and Get the Party Started by Pink. A great group of people showed up, several actually paid attention :-). Here's what we drank: 67 mixed drinks 70 glasses of wine 17 domestic beers 20 imported beers 40 micro-brews 27 soft drinks 5 bottles of water

Disclaimer and Copyright

In this podcast we talk about:

• Flat file to distributed relational with application rewrite
• Flat file to relational with no application rewrite the right way
• VSAM and IMS to relation with no application rewritfe the wrong way
• SQL Server to DB2 stored procedures
• Migrating to a new application using data sharing

05/01/2006 - Series 1 Podcast 13 SQL Coding (back to series 1), (back to top)

Notes by Dan....Since all our podcast equipment got (allegedly) stolen we had to practice with the new equipment. Our loss, your gain!! We're once again in Baltimore. Sue is drinking a red wine and I'm have Sierra Nevada Pale Ale. We talk more about SQL performance, a topic that we can talk about forever. Surprise! I provide the musical beds including Mexico by Cake, Rock and Roll Jewel Intuit by Danny P, and Pink Houses by John Mellancamp.

Disclaimer and Copyright

In this podcast we talk about:

• Stage 1, stage 2, stage 3 filtering
• fat versus skinny cursors
• programmatic joins versus SQL joins
• Table join sequence

04/07/2006 - Series 1 Podcast 12 SQL Coding (back to series 1), (back to top)

Notes by Dan....What's wrong with me? I get the volume wrong on the songs, continually push the wrong buttons, start reading the wrong commercial, and then ultimately accidentally leave all of the podcast equipment (several hundred dollars worth) at the airport (presumed stolen by Hertz employees). I'm a mess. In spite of these shortcomings Susan and I have delivered some more great tips and solid information in this podcast about coding SQL for performance. We've only scratched the surface as you can talk about this topic for days or longer. We may have to do another one! Once again we are in Baltimore, our home away from home and one of our favorite cities. Susan drank a BV Coastal White Zinfadel and then a Merlot, and I had Stoli Vodka on the rocks. The music beds were Excitable Boy by Warren Zevon, Hit It by The Offspring, and All the Young Dudes by Mott The Hoople. Dig it, baby!

Disclaimer and Copyright

In this podcast we talk about:

• Basic Recommendations for SQL Performance
• Ordering and Sorting
• What Happens When Nested Table Expressions are Merged
• Ordering of Predicates
• Why Put Complexity in SQL and When is it a Performance Advantage

03/09/2006 - Series 1 Podcast 11 Special Anniversary Podcast - Tuning SQL; Stats and Optimization (back to series 1), (back to top)
Notes by Dan....SQL optimization is extremely important, and can get complicated. However, for most of us the job is easy. That's because the DB2 optimizer does all the hard work for you, and the vast majority of the time gets that job done correctly. The optimizer, linke any person, needs information to get the job done right, and this is where you need to focus the majority of your effort. Collecting and understanding statistics is important to doing a good job. Understanding how the optimizer uses statistics to influence access path can help make your job easier. In this podcast we talk about how the optimizer uses statistics, and how the DBA needs to understand this, collect the proper statistics at the right time, and test and implement access path changes. Dan is drinking a St. Pauly Girl and a Sam Adams, and Susan is having a BV Coastal Shiraz. Given it's our one year podcast anniversay we are particularly happy, and playing such celebratory songs as All I wanna Do by Sheryl Crow, My Hero by the Foo Fighters, and Can't Get Enough of You Baby by Smashmouth. Pop a cork for one year of fun and DB2!

Disclaimer and Copyright

In this podcast we talk about:

• The DB2 optimizer
• Statistics
• Column correlation
• Runstats
• Rebinding

02/13/2006 - Series 1 Podcast 10 Making Love (back to series 1), (back to top)

Notes by Dan....How do we possibly mix the concept of DB2 and making love? It's easy! You see, there are many aspects to a job working with DB2, and there are many people you have to interact with to do your job really well. We take our years of real world experience, and stories of the people we've worked with to help you understand who you need to work with, and what you need to understand. Newer technologies available today bring people that were traditionally not involved with DB2 into the fold so there are new people you have to make friends with to be really effective and take advantage of this technology! We started with what we thought was our first non-technical podcast, but of course we got technical...we just can't help ourselves! Sue is drinking BV Coastal Merlot, and I'm drinking some Kettle One vodka. The song beds include Love You Madly by Cake, I think I Love You by the Partridge Family, and Bursting with Love by Randy Herman. Snuggle up with a friend and a beverage and injoy a little DB2 lovin'!

Disclaimer and Copyright

In this podcast we talk about:

• Database Design
• Performance Tuning
• Memory and Version 8
• Backup and Recovery
• New Technologies

12/14/2005 - Series 1 Podcast 9 Minimizing Change in DB2 (back to series 1), (back to top)
Notes by Dan....we all fear change, and as DBA's we should fear change to our data. OK, just kidding! In a high performance design we need to consider the details, and one of those details is knowing what data changes and how often. When this is understood then tables can be designed to help avoid the impact of change. This impact is the amount of logging the database performs, and the amount of table maintenance the DBA has to perform. This podcast once again brings us back to our home away from home in Baltimore. We're both drinking some (not so good) Nissley 2004 Chambourcin from Pennsylvania. It's our special Christmas and holiday show, so we've included some very special holiday music and good cheer. This is the first podcast on my new PC (I'm putting the old girl to rest after four and a half years), and the sound is a bit better. The special songs include a version of Do They Know It's Christmas by The Barenaked Ladies, a histerical organ version of Rudolph The Red Nosed Raindeer also by The Barenaked Ladies, and Merry Christmas I Don't Want to Fight Tonight by The Ramones. Grab a glass of holiday cheer and enjoy!

Disclaimer and Copyright

In this podcast we talk about:

• Data capture
• Copy yes for indexes
• Table design to minimize logging
• How DB2 helps to minimize logging
• Detecting change
• Minimizing change and table maintenance
• Optimizing blind updates

10/26/2005 - Series 1 Podcast 8 The Evolution of the DBA (back to series 1), (back to top)

Notes by Dan....The features, operation, accessability, and related components of the database have grown over the years. The role of the DBA in this new world is more complicated than ever, and yet DBAs are finding themselves more often than not in a situation that is out of their control and without the knowledge to support what is going on. In this very special podcast we talk about how the DBA has evolved, where the new responsibilities and challenges lie, and give some advice as to how to meet some of these challenges. This podcast finds us in Berlin, Germany, and is our first ever podcast with special guests. We're joined here by our very good friends Steve McDougall and Kurt Stuyf who, like us, work with DB2 every single day in large production implementations. I'm drinking a Rodenbach (a present from Kurt), Kurt and Steve are drinking Coors Lite (a present from Dan), and Susan is drinking Mongozo Banana beer (another gift from Kurt). Since this was our first group podcast I didn't get the sound just right, and it's a little distorted, but the valuable lessons learned are still well worth the listen. Also, don't be fooled by the introduction when I say it's podcast 7...I goofed. Then again, as always the podcast is unscripted and unedited so what you get is genuine advice from people really doing the work. Grab yourself a brew and join the party! Oh, and for you DBAs out there tell your managers about this one so they'll understand how hard you work! I forgot to take photos at the podcast so I've included photos of us with Kurt and Steve, along with a couple of photos we took at a crazy cafe called the KLO in Berlin....can you believe what we're drinking out of!? The song beds include You Get What You Give by New Radicals, Take Your Mama by Scissor Sisters, and Bright Future in Sales by Fountains of Wayne. Finally, on the podcast, towards the end...have I been drinking to much, or do I just start naturally trying to sound like Kurt?

Disclaimer and Copyright

In this podcast we talk about:

• What makes a great DBA
• Changing Roles of DBAs
• Communication
• New Technologies
• Tuning efforts

10/07/2005 - Series 1 Podcast 7 DB2 UDB for z/OS Version 8 (back to series 1), (back to top)

Notes by Dan....DB2 for z/OS version 8 is the largest release in DB2 history. A release this significant requires careful consideration, and a very thorough migration plan. Come and listen to our best podcast ever! Broadcast from Baltimore, completely unscripted and unedited, you can hear our analysis and opinions on this significant release of version 8. We give planning advice, talk about features, and share some of our real world experience with this release! I'm drinking Amstel Light, and Susan (although she hates to admit it) is drinking a glass of Beringer White Zinfidel. Susan takes a shot at her first live read commercial, and the music beds include Welcome to the Jungle by Guns and Roses, Take a Chance on Me by ABBA, and Here I Go Again by Whitesnake. Grab a brew, relax, and enjoy!

Disclaimer and Copyright

In this podcast we talk about:

• Major V8 features
• Modes of Operation
• SQL Optimization and Statistics
• DFSORT
• Vendor Products and Application Testing
• Our Experience with Performance, Stability, and Functionality

08/23/2005 - Series 1 Podcast 6 Inserts (back to series 1), (back to top)

Notes by Dan....Inserts can be the number one performance concern for large high volume applications. Proper understanding of your input data, cluster, and free space are all critical in acheiving the desired level of insert performance. Also, you need to have an understanding of how inserts work to build a better design and reorg strategy. This podcast finds us near Johannesburg, South Africa! We are both drinking Spier Cabernet from the Stelenbosch wine region near Capetown. The song beds include The Pump It by the Black Eyed Peas, and Everything's Not Lost by Coldplay.

Disclaimer and Copyright

In this podcast we talk about:

• Freepage and Pctfree
• clustering
• Sequential processing
• Member Cluster
• Page Splitting
• Insert Algorithms

07/14/2005 - Series 1 Podcast 5 Clustering (back to series 1), (back to top)

Notes by Dan....Clustering is often ignored during application development and performance tuning. Clustering in general gives us the ability to order data in our tables in a given sequence, and is useful for sequential processing, joins, and data retrieval by a common key. It can also influence table organization and insert performance. Take a listen to this podcast on the many influences and effects of clustering. This podcast brings us to Parkersburg, West Virginia. We are both drinking Canoe Ridge Merlot. The song beds include The Keys to Her Ferrari by Thomas Dolby, and Evangeline by Mathew Sweet.

Disclaimer and Copyright

In this podcast we talk about:

• What is clustering
• The clustering index
• Sequential processing
• List prefetch and index lookaside
• Clustering and joins
• Multi-Dimensional clustering

06/02/2005 - Series 1 Podcast 4 Dynamic SQL (back to series 1), (back to top)
Notes by Dan....dynamic SQL is becoming more prevalent in our online applications due to the increase in remote access, web interfaces, Websphere, etc. This presents special challenges in the area of predicting dynamic SQL performance, monitoring, and controlling dynamic SQL. So, come here us drink and discuss options for monitoring and controlling dynamic SQL in DB2 for z/OS, from version 6 through version 8. This podcast brings us back to Baltimore, Maryland (our home away from home). We are both drinking BV Coastal Cabernet. The songs we dropped in are Never Gonna Stop Me by Rob Zombie, and Lets Go by The Cars.

In this podcast we discuss a product called Dynamic SQL Cache Control. For more information about this product, please contact us:

Disclaimer and Copyright

In this podcast we talk about:

• Why dynamic SQL
• The monitoring challenges and solutions
• Static Profiling
• Dynamic SQL Cache Control

05/04/2005 - Series 1 Podcast 3 Sequential Numbering (back to series 1), (back to top)
Notes by Dan....Sequential numbering can be one of the biggest bottlenecks in an application if improperly implmented. DB2 does offer some solutions to the problem of sequential numbering, but you have to be aware of some of the issues that surround these features. So, come here us drink and discuss the best way to do sequential numbering in DB2, from version 6 through version 8. This podcast brings us back to Baltimore, Maryland (our home away from home). Susan is drinking some cheap red wine from the lounge, and then a Bud Light. I'm drinking Amstel Light. I did all the voice-overs live this time. The songs we dropped in are Buddy Holly by Weezer, and Do You Realize? by The Flaming Lips. Enjoy!
Disclaimer and Copyright

In this podcast we talk about:

• Next key tables
• Identity columns
• The best way to use identity values
• Sequence objects

03/25/2005 - Series 1 Podcast 2 Availability (back to series 1), (back to top)
Notes by Dan....DB2 V8 includes many new feaures for availability, and so we thought we'd address those features, but also talk about how application and database design are a very big part of high availability. This podcast was recorded at Susan's home in Springfield, Illinois. Susan was having some Trapiche Cabernet from Chile, and I had a couple of Sierra Nevada Pale Ale. Drop-ins include Opera Singer by Cake, and Goody Two Shoes by Adam Ant. OK, so now we're a bit looser...maybe too loose! There are a few silent moments as I forget what I'm doing, and keep pushing the wrong buttons on the computer, so keep listening through those parts (we don't edit).
Disclaimer and Copyright

In this podcast we talk about:

• Availability Features in V8
• Careful Usage of New V8 Features
• Designing a database for high availability
• Designing an application for high availability

03/03/2005 - Series 1 Podcast 1 DPSI's (back to series 1), (back to top)
Notes by Dan....Here we go! This is our very first podcast, and as far as we know the first DB2 podcast ever. We chose the V8 topic of DPSI's because they are probably asked about more than any other V8 feature, and misunderstood by many. This podcast was recorded high atop the Marriott Waterfront in beautiful Baltimore, Maryland. Susan was having some Robert Mondavi Merlot, and I had a couple of Amstel Lights. We also had some pretzels. Drop-ins include Rock the Nation by Montrose, and Fly by Sugar Ray. I was going to drop in Goody Two Shoes by Adam Ant, but forgot to...maybe next time! I was pretty excited and I think I drank most of my beer in the first one minute. This being our first one we may be a bit stiff, but we'll get the hang of it! We hope you like it, and please give us some feedback and topic suggestions!
Disclaimer and Copyright

In this podcast we talk about:

• Current issues with NPSI's
• Alternatives to NPSI's
• Challenges with UNION in view subquery pruning
• Discussion on V8 DPSI's
• Challenges with DPSI's
• Partition elimination issues

We're using Dan's PC, an HP Pavilion dv1010us, and the following sweet stuff:

• CAD Equitek E100 supercardiod large diaphram side address condenser microphone (presumed stolen by Hertz employees)
• Behringer MIC100 tube ultragain preamp (presumed stolen by Hertz employees)
• Pinnacle table top mic stand (confiscated by airport security after going through 6 times!)
• QuikLok A188 Desktop Tripod Microphone Stand (presumed stolen by Hertz employees)
• CAD Equitek E100**2 supercardiod large diaphragm side address condenser microphone
• QuikLok A188 Desktop Tripod Microphone Stand
• Behringer MIC200 tube ultragain preamp
• Mixcast Live podcasting software
• Dell Inspiron 8100 (retired after series 1, podcast 8)