This post is about the IBM Db2 Analytics Accelerator, sometimes (and hereinafter) referred to as IDAA.
First of all, for those who don’t know, let’s start with what it is. IDAA is a high-performance component, typically delivered as an appliance, that is tightly integrated with Db2 for z/OS. It delivers high-speed processing for complex Db2 queries to support business-critical reporting and analytic workloads.  

The general idea is to enable HTAP (Hybrid Transaction Analytical Processing) from the same database, on Db2 for z/OS. IDAA stores data in a columnar format that is ideal for speeding up complex queries – sometimes by orders of magnitude.

Now there is a lot more to IDAA, but we won’t cover it here in today’s blog. If you want more details, I direct you to the following links:

Anyway, the real purpose of today’s blog entry is to alert IDAA users that you need to be aware of some recent and upcoming support and version issues.

IDAA Version 7

The current version of IDAA is V7.5; it was announced October 15, 2019 and released for GA December 6, 2019. But many customers are not there yet. This is not surprising given that it has only been about 4 or 5 months since it has become available. Nevertheless, it offers an abundance of great functionality and usability improvements. At the top of the list are greater scalability and improved synchronization.

Because the data in an IDAA is stored separately from the data in the primary Db2 for z/OS system, when the data is changed in Db2 for z/OS it must be migrated to the IDAA. This causes latency, where the data differs between the two systems. Of course, this is not ideal.

Well, the latest and greatest iteration of IDAA has greatly improved things with Integrated Synchronization, which provides low-latency data coherency. Db2 12 for z/OS (FL 500) delivers the Log Data Provider, which to capture changes and funnel them to IDAA. It is quick, uses very little CPU, and is zIIP-enabled. This greatly improves the latency between Db2 for z/OS data and IDAA data, to the point of it becoming mostly irrelevant.

Additionally, V7 was the first version of IDAA to allow deployment on IFLs, instead of on a separate physical piece of hardware. This means you can accelerate Db2 for z/OS queries completely on the mainframe. And V7.5 expands the scalability of IFLs.

Important Information for Laggards

Perhaps the most important piece of information in today’s blog post though is for those of you who are still running older versions of IDAA… specifically, V4. The end of service date for IDAA V4 is imminent – April 30, 2020 – and there will be no extension of this date. So if you are still on V4, it is time to upgrade!

Fortunately, you can upgrade to IDAA V5 at no cost. Sure, V5 is not the most current version of IDAA, but IBM has not issued an end of service (EOS) date for it yet. The probable EOS date is tentatively set for the first half of 2023 (which is the same for the IBM PureData System for Analytics N3001 on which this earlier IDAA is based.

Today’s Bottom IDAA Line

If you are looking for an efficient, cost-effective query accelerator for your complex Db2 queries you should look into IDAA V7.5.

And if you are still running V4, update soon (by the end of the month?) to avoid running on an out of service version of IDAA.