April 30, 2012 is coming closer, now what to do?

Unless you believe the Mayan calendar to be right and you are still on DB2 8 for z/OS, the time to act is now.  DB2 version 8 for z/OS end of support happens on April 30th 2012.  So now is a good time to look at the past, present and future.

The past

DB2 version 8 for z/OS: do you remember first hearing about it? I do. Version 8 was announced with great new features.  For the first time we would be using 64bit addressing, doing away with quite a bit of the 2Gb restrictions many of us were facing. That was the good news.  The “bad news” was that during the migration, we would not be able to fallback to DB2 version 7.  Fear struck in our mainframe hearts, it took the poor folks of IBM a long time to convince us that it was going to be all right. Most of us have made that leap (some sooner than others), to find out it wasn’t such a big deal anyway.

This migration discussion sometimes drew attention away from the really great application and administrative features that were new to version 8.  Although I don’t want to bore you with a full list of version 8 features, I do want to remind you of three features, I believe beneficial to any site.


First, multi-row fetch/insert/delete in batch processing. You might be running version 8 load modules, but if you haven’t introduced multi-row processing in your batch programs, you are missing out on an amazing feature and I would almost say you’re not using version 8.

Secondly, Real Time Statistics (RTS), although technically an APAR’ed version 7 feature, many of us didn’t hear about it until version 8 GA time. RTS should reduce the time spent on maintenance. As many of us are faced with ever smaller DBA teams, I don’t see how you could do without RTS.

Thirdly, table controlled partitioning. I know there have been many discussions on this subject, but any feature that allows me to potentially get rid of an index and leaves me with more design options, is essential in my book.

The present

Most of us are currently on version 9, and the number of DB2 10 customers is rapidly increasing. Regardless of if you are currently still on version 8 or version 9, you should start making plans to migrate to version 10.  DB2 10 for z/OS is a very stable release, with “out of the box” CPU savings and it should be your goal. This is easy to explain when you’re on DB2 version 8. You’re facing end of support and IBM allows a skip migration into DB2 10. You will get nice CPU savings and you avoid doing two migrations.  Less downtime and less weekend and night work.

If you have already migrated to DB2 9, you might be reluctant to start thinking about migrating to DB2 10. I completely understand that feeling, but it is pure and simple math. At the beginning of February, IBM announced “end of support” for DB2 9 to be June 27, 2014. You might think that this is still a far distant future, but allow me to give you the following math example:

Considering that migrating to DB2 10 will again be a 3-phase (CM-ENFM-NFM) process, it will take some time to rollout all of your environments to NFM.  Currently, I find most sites to have a rollout period of 12 - 18 months with a new DB2 release.  That means that if you’re on the high end of this spectrum, you should start your migration work on January 27, 2013.  That is roughly 9 months from now.  Are you familiar with the pre-requisites?  If you are a site that has been avoiding using DFSMS, you will have some work to do before you can migrate to version 10.  Is your CICS, IMS or Websphere at the correct level? I know this is not the case at certain sites, but that means quite a bit of work still needs to be done before you can actually start your DB2 migration. So at least verifying where you are, what you are still needing to do and making a realistic planning is something you want to look into - now!

The future

It’s quite clear that when IBM announces the end of support of DB2 9, they are working hard on the next release of DB2.  No doubt it will hold nice new features and improvements. However, you shouldn’t wait for the next release as the future is here with DB2 10. At least it would be, if we were living in a temporal world. Fortunately for us, our data has the capability to exist in a temporal design.  Our queries can do nothing short of time travelling.  How cool is that?  The only time machine you need is called DB2 10.

DB2 10 is definitely a version that will set us up for the future. The XML-support is excellent, even allowing you to add a validation schema during the table definition, much like a check constraint.  We now have full support of Xpath, SQL/XML and (drum roll please) Xquery.  Since October 2011, it has been possible to run Xquery against DB2 10 for z/OS.  It does require you to be up to date on maintenance.

DB2 10 will satisfy your auditors as we now have the capability to separate responsibilities.  This feature requires nothing but the database engine as it comes out of the box.  It has SECADM, SYSADM without data access and other great security possibilities built in.

Beyond the technical aspects, DB2 10 is also a release for the future from an executive level point of view. Through its CPU savings right out of the box, it is helping to lower the “alleged” high cost of mainframes.

With these and other great benefits and features, and with knowing how long your migration takes, start counting backwards and let April 30, 2012 be the day that you decided it’s time to reflect on the past, live in the present and start thinking about the future.

Kurt Struyf, Suadasoft

IDUG content committee

Recent Stories
The IDUG Board of Directors Slate for 2017-2018 is Open for Voting

Open Board of Directors Position - Apply Before the January 13, 2017 Deadline

Enable Social Login