DB2DD: Roger, thanks very much for talking with us. How did you end up in DB2?
Roger: I had worked in simulation applications and systems programming at other companies for about ten years, before a friend of mine got me an interview at IBM. I've been a part of the DB2 team for the past 23 years. Chief packrat, a.k.a. historian, should be part of my job description. In that time, I've worked on everything from installation to precompiler to authorization. I am still digging in the code for precompiler and install regularly.
DB2DD: Well, you obviously do a lot of stuff. What is your job description?
Roger: Well, my title is "strategist," but the real title should be, "whatever it takes."
DB2DD: Then maybe it would be easier for you to describe for us a typical day for you.
Roger: I like to say that I work half days - from 7 to 7. A classic day for me is from the silicon to the sky. I keep an eye on things from a business side, doing licensing forecasts and the like. We're doing pretty well by the way, up 13% in revenue year on year for 2001 and the momentum is continuing in the first quarter. I design and communicate the strategy to the team and to customers. I consult with customers a lot--about 50% of my time is dealing with customers. I hang out a lot on the DB2 for z/OS support page and on the DB2-L listserver . I review code and manuals, I write three or four columns a year for the IDUG Solutions Journal , and I still teach a DB2 installation class. I do around 100 presentations a year, and I typically do not repeat the same one twice. I spend a lot of time in the air. I flew 50,000 miles just in the 4th quarter last year.
You could summarize by saying that my job is whatever it takes that day to help customers and IBM win.
DB2DD: What is your favorite part of your job?
Roger: The absolute best part of the job is getting a new function to work. YES!!! There's nothing like that feeling of getting a new DB2 function to work. I love winning. I love it when a customer buys DB2 over a competitor's database product.
DB2DD: What is the strangest question you ever got from a customer?
Roger: [laughs] You mean today? I've gotten some that I don't want to remember. I do answer a lot of customer requirements and some of the requirements basically ask us to improve upon the speed of light.
DB2DD: The Data Management Technical Conference is traditionally a time when lots of good new information gets revealed. What are you going to talk about at the conference this year?
Roger: The big topic is a technical two-parter with Curt Cotner on a technical introduction to the next version of DB2 for z/OS.
DB2DD: That sounds interesting. Can you give us a sneak preview?
Roger: It's a big release. We expect our next delivery to make a fundamental change in many areas, reengineering much of DB2. There are key improvements that help improve scalability, improved ability to port applications, much better ability to deliver continuous availability. The ability to handle very large databases has fundamental changes. Support for key vendor applications is compelling. The enhancements for 64-bit virtual storage will make management simpler, and improve scalability and availability. This new version will break through many old limits.
DB2DD: Well, let's look at some of these things a little deeper. For example, what kind of reengineering are you talking about?
Roger: I like to say that DB2 is going through a renaissance in the rebuilding we've been doing. For example, the amount of change for the index is almost like creating a type 3 index, only we don't call it that. We will be able to use indexes more effectively, reducing the space in variable-length indexes, being able to have index-only access with variable-length data and being able to use the index when the predicates do not match.
We are also planning on allowing partitioning and clustering to be independent. What is the importance of that? Some people would like to partition without an index and be able to cluster on any index. It means in some cases people can live with one less index and reduce the occurrences of "death by random I/O" by taking advantage of sequential I/O. This will improve INSERT, DELETE, LOAD, REORG and UPDATE processing.
We have more flexibility in indexes with longer index keys, up to 2000 bytes, which is something we needed to do for Unicode.
DB2DD: Are there a lot of changes for Unicode coming in the next release?
Roger: Well, you could certainly see it coming. For all of our customers, it's not a question of if they are moving to Unicode, it's when . Multinational companies need it. The Java character set is Unicode.
Version 7 of DB2 for z/OS and OS/390 ® delivered support for Unicode-encoded data. The encoding scheme can represent the code points of many different geographies and languages. You can easily store multilingual data within the same table or on the same DB2 subsystem.
In the next version, we plan to make Unicode support more flexible, with such enhancements as the ability to join a Unicode table to one that is ASCII or EBCDIC. SQL statements and literals can be Unicode or EBCDIC. Many of the DB2 catalog character columns will be converted to Unicode, so Unicode is for everyone.
DB2DD: DB2 for z/OS is known for its robust availability features. What are some of the things users can look forward to in that area?
Roger: A huge enhancement in this area is schema evolution. For some changes to tables or columns today, our users have to drop an object and recreate it, which in some unfortunate cases can take hours or days. In the next version, you will be able to make more changes using an alter in a fraction of a second. For example, you will be able to extend numeric and character columns, you will be able to change between CHAR and VARCHAR. You will also be able to add a new partition to an existing partitioned table space. You can rotate the partitions, which will allow you to do such things as keep rolling through the most current 36 months of data.
DB2DD: You mentioned very large databases as something that the next version is addressing. What can people look forward to there?
Roger: The biggest impact of the z/Architecture™ on DB2 is the ability to have large real memory support. Prior to the zSeries™, customers were limited to 2 GB real storage due to the 31-bit addressing of the S/390 architecture. The real storage limit of 2 GB is a leading performance inhibitor for many high-end customers. Another performance inhibitor is the 2 GB virtual storage limit for the main DB2 (DBM1) address space.
In DB2 Version 3 we provided hiperpools to offer some relief, but many customers need more. Version 6 allowed customers to use large real storage by moving to buffer pools in data spaces.
IBM plans to deliver 64-bit virtual storage addressing for the DB2 for z/OS product in a future release. Instead of hiperspaces or data spaces, the single large address space can allow easier management of storage as we transition from multiple 31-bit address spaces to a few 64-bit address spaces, improving both availability and scalability. Think about it! In a single address space, we have addressability up to 16 exabytes. Right now, we don't know anyone who has that big a database. But we'll be there when there are.
DB2DD: Is there any one thing in the next release of DB2 that you think is the most exciting?
Roger: The most exciting feature really depends on the customer. This version is about breaking through limits, and our customers face many different limits: scalability, continuous availability, cross-platform compatibility, and usability are some of the "ities." Also, anyone who is interested in enterprise resource management products like SAP, PeopleSoft, and Siebel should understand that many of these enhancements will help with those large applications, too, such as larger memory, Unicode, and SQL flexibility improvements. I'll let you talk to Curt Cotner about more details for SQL and networking enhancements, but I will say that we plan to have longer table and column names and we will be raising several limits partly as a result of this. We are also planning on supporting multi-row fetch and insert, which these ERM applications use frequently.
People should make a point of coming to the Data Management Technical Conference to get more details on all the great stuff that is being planned for Vnext.
DB2DD: Roger, I appreciate your taking the time to talk to us. What are you looking forward to next?
Roger: I am really looking forward to going to Yosemite with my wife to celebrate our 30th wedding anniversary.
Roger L. Miller is the lead DB2 for z/OS strategist, but he started in DB2 as a programmer. He works on product strategy, planning, development, writing, and helping customers to use the product with a focus on performance, availability, and security. He can be reached at millerrl@us.ibm.com.
Comments (Undergoing maintenance)
Table of contents
Next steps from IBM
My developerWorks community
Tags
Use the slider bar to see more or fewer tags.
Popular tags shows the top tags for this particular content zone (for example, Java technology, Linux, WebSphere).
My tags shows your tags for this particular content zone (for example, Java technology, Linux, WebSphere).
Popular article tags |
My article tagsSkip to tags list
Popular article tags |
My article tags
