Thursday, December 17, 2020

Db2 Utilities and Modern Data Management

Db2 utilities are the unappreciated, and often over-looked, workhorses of your mainframe Db2 environment. They perform the dirty work that has to be done to populate, organize, backup, and recover your vital mainframe data. Without them, building effective Db2 databases, managing data, optimizing performance, and even accessing mainframe data would be a lot more difficult than it currently is.

The Situation 
Think about the Db2 utility situation at your shop. If you are like most organizations you will have Db2 utilities regularly running all the time. There are load and unload tasks running to refresh data for development and testing, for moving data between environments for analysis and processing, and for various other purposes. The LOAD and UNLOAD utilities bear a lot of the hard work for data movement. 

You are also most likely reorganizing data using a REORG utility for most of your Db2 table spaces and probably indexes, too. In many cases reorganization jobs are scheduled to run on a regular basis: weekly, monthly, quarterly, etc. Frequently you just set these jobs up when the object is created. The job gets scheduled and is just run without anybody taking a look at them unless, or until there are performance problems. 

Then there are COPY and RECOVER utilities for backing up and recovering data when there are problems. The image copy backup jobs are running all the time, taking either full or incremental copies to ensure that you can recover data in case problems are encountered. The copies are running all the time, but the recover jobs (hopefully) are not running all the time! 

You are also going to be running the RUNSTATS utility to gather statistics for Db2 to use for query optimization. Depending on how often your data changes, you may be running RUNSTATS frequently or infrequently. Many times the same fate as REORG befalls RUNSTATS… that is, it is scheduled and forgotten about unless problems arise. 

There are other utilities, like CHECK which is used to verify the integrity of data. You are probably not running this one very often but when you need it you want it to run fast, right? 

So, all of these utilities are “out there” running and consuming CPU to move, copy, and manage your Db2 data. But are they being run effectively as possible? 

Moving to the Modern Db2 Utility Way 
I think by this point everybody will agree that utility type processing is not just critical, but mandatory for a Db2 environment. But just running with the bare basics is not the best approach. 

If we think about data movement with unload and load processing there are several things that you might want to consider for improvement. First of all, consider the speed and performance of the unload and load tasks. You probably want these jobs to run as fast as possible – that is, to consume as little elapsed time as possible to complete. After all, you are probably using these utilities to build environments or even refresh portions of an environment… and there will be developers and testers waiting to use that data as soon as it is available. Using the fastest utility programs available will minimize the wait time and make your developers and testers more productive. Furthermore, you want these tasks to consume as little CPU as possible to reduce your monthly mainframe bills! 

In some cases you might want to re-consider unloading and loading altogether, using alternate utilities and offerings that can clone an entire subsystem or move data outside the control of Db2 at the data set level. 

If we think about reorganization, it is likely that you are running REORG tasks that don’t need to be run, at least not as regularly as they are being run. At the same time, it is also likely that you are not running other REORG tasks as frequently as you should, thereby causing every other task that accessing the data to degrade. Fortunately, you can use RTS (real time statistics) to help guide when you should (and should not) reorganize your data. In the best case the utility itself relies on RTS to figure out if it needs to run and runs when it makes sense only. Failing this, you are again likely consuming more CPU than is necessary (either running unneeded REORGs or accessing poorly organized data, as the case may be). 

If you think about your backup and recovery situation, the issue is likely complexity. Sure you want COPY and RECOVER utilities that run fast and consume minimal CPU, but the big issue is analysis. By that I mean, when you need to recover you want to make sure that you can use the image copies (and, of course, the log) to recover and meet your RTOs (recovery time objectives). But creating recover jobs on-the-fly, in a probably complicated environment with inter-related tables and data, can be difficult. And doing so when there is an outage, which is usually the case, exacerbates the situation. Using intelligent utilities to create the right image copies and to automatically build an appropriate recovery strategy when needed should be the modern approach.

And not to neglect RUNSTATS and CHECK, you want both of those utilities to run as fast as possible, consuming minimal CPU, too. And you want guidance on when and how to run them using available RTS, statistics, and any system information available. 

What Can You Do? 
One approach is to use modern utilities, not only built for speed but that incorporate AI and machine learning to automate and improve the Db2 utility experience. BMC Software is once again on the vanguard with its BMC AMI utilities for Db2

The first question you probably have is "What the heck is AMI?" Well, AMI, which stands for Automated Mainframe Intelligence, is technology that is being infused into BMC’s product line to leverage AI, machine learning, and predictive analytics to achieve a self-managing mainframe. 

BMC AMI Utilities for Db2 are designed for modern complex Db2 environments. They use a centralized, intelligent architecture (see diagram below) designed specifically to handle the complexity facing IT today. Through intelligent policy-driven automation, you can use the AMI Utilities for Db2 to manage growing amounts of data with ease and, at the same time, deliver full application availability. 

Figure 1. BMC AMI Utilities for Db2

If you are looking to reduce CPU and elapsed time by as much as 75%, eliminate downtime while delivering full application availability, lower disk usage, eliminate sort in your REORGs, and simplify complex utility operations, then it makes sense to take a look at the BMC AMI Utilities for Db2. 


You might also want to take a look at this blog post from BMC that discusses how to Save Time and Money with Updated Unload Times 

And this analysis of the BMC next generation REORG technology from Ptak Associates

Wednesday, November 18, 2020

Deleting "n" Rows From a Db2 Table

I regularly receive database- and Db2-related questions via e-mail. And that is great, but I don't always get a chance to respond to everything. If you've sent me a question and I haven't replied, I apologize. But every now and then, I will use one of the e-mail questions in my in-box and write about it in the blog. Today's question is this:

How do you delete N rows from a Db2 table?

Also, how do you retrieve bottom N rows from a Db2 table without sorting the table on key?

And here is my response:

First things first, you need to refresh your knowledge of "relational" database systems and Db2. There really is no such thing as the "top" or "bottom" N rows in a table. Tables are sets of data that have no inherent logical order.

With regard to the result set though, there is a top and a bottom. You can use the FETCH FIRST N ROWS ONLY clause to retrieve only the first N rows, but to retrieve only the bottom N rows is a bit more difficult. For that, you would have to use scrollable cursors.

A scrollable cursor allows you to move back and forth through the results set without first having to read/retrieve all of the rows before. I suggest that you read up on scrollable cursors in the Db2 SQL Reference manual and the Db2 Application Programming manual. All Db2 manuals can be downloaded in Adobe PDF format for free over the IBM web site.

Basically, you would want to FETCH LAST from the scrollable cursor and then loop through with a FETCH PRIOR statement executing the loop N-1 times. That would give you the "bottom" N of any results set -- sorted or not.

As for your other question, I am confused as to why you would want to delete N rows from a table. Doesn't it matter what the data in the rows is? My guess is that you are asking how you would limit a DELETE to a subset of the rows that would apply to the WHERE condition of the DELETE. The answer is, you cannot, at least not without writing some code.

You would have to open a cursor with the same WHERE conditions specifying FOR UPDATE OF. Then you would FETCH and DELETE WHERE CURRENT OF cursor for that row in a loop that occurs N times. Of course, that means you have to write a program to embed that SQL in.

Hope this answer helps...

Wednesday, October 21, 2020

Automation and the Future of Modern Db2 Data Management

Recently I was invited by BMC Software to participate in their AMI Z Talk podcast series to talk about modern data management for Db2... and I was happy to accept.

Anne Hoelscher, Director of R+D for BMC's Db2 solutions, and I spent about 30 minutes discussing modern data management, the need for intelligent automation, DevOps, the cloud, and how organizations can achieve greater availability, resiliency, and agility managing their mainframe Db2 environment.

Here's a link to the podcast that you can play right here in the blog!

Modern data management, to me, means flexibility, adaptability, and working in an integrated way with a team. Today’s data professionals have to move faster and more nimbly than ever before. This has given rise to agile development and DevOps - and, as such, modern DBAs participate in development teams. And DBA tasks and procedures are integrated into the DevOps pipeline. 

And as all of this DevOps adoption is happening, the amount of data we store, and have to manage, continues to grow faster than ever before.

These are just some of the challenges that Anne and I discuss in this podcast... and at the end, Anne even asks me to predict the future... 

I hope you'll take the time to listen to our discussion and sharing your thoughts and issues regarding the resiliency and agility required to succeed with modern data management and Db2 for z/OS.


I’d also like to extend an offer to all the listeners of this BMC podcast (and readers of this blog post) to get a discount on my latest book, A Guide to Db2 Performance for Application Developers. The link is

There’s also a link to the book publisher on home page of my website. Once you are there, click on the link/banner for the book and when you order from the publisher you can use the discount code 10percent to get 10% off your order of the print or ebook.


Monday, October 19, 2020

Improving Mainframe Performance with In-Memory Techniques

 A recent, recurring theme of my blog posts has been the advancement of in-memory processing to improve the performance of database access and application execution. I wrote an in-depth blog post, The Benefits of In-Memory Processing, back in September 2020, and I definitely recommend you take a moment or two to read through that to understand the various ways that processing data in-memory can provide significant optimization.

There are multiple different ways to incorporate in-memory techniques into your systems ranging from system caching to in-memory tables to in-memory database systems and beyond. These techniques are gaining traction and being adopted at increasingly higher rates because they deliver better performance and better transaction throughput.

Processing in-memory instead of on disk can have a measurable impact on not just the performance of you mainframe applications and systems, but also on your monthly software bill. If you reduce the time it takes to process your mainframe workload by more effectively using memory, you can reduce the number of MSUs you consume to process your mission-critical applications. And depending upon the type of mainframe pricing model you deploy you can either be saving now or be planning to save in the future as you move to Tailored-Fit Pricing.

So it makes sense for organizations to look for ways to adopt in-memory techniques. With that in mind, I recommend that you plan to attend this upcoming IBM Systems webinar titled The benefits and growth of in-memory database and data processing to be held Tuesday, October 27, 2020 at 12:00 PM CDT.

This presentation features two great speakers: Nathan Brice, Program Director at IBM for IBM Z AIOps, and Larry Strickland, Chief Product Officer at DataKinetics.

In this webinar Nathan and Larry will take a look at the industry trends moving to in-memory, help to explain why in-memory is gaining traction, and review some examples of in-memory databases and alternate in-memory techniques that can deliver rapid transaction throughput. And they’ll also look at the latest Db2 for z/OS features like FTBs, contiguous buffer pools, fast insert and more that have caused analysts to call Db2 an in-memory database system.

Don’t miss this great session if you are at all interested in better performance, Db2’s in-memory capabilities, and a discussion of other tools that can aid you in adopting an in-memory approach to data processing.

Register today by clicking here!

Wednesday, October 14, 2020

Db2 12 for z/OS Function Level 508

This month, October 2020, IBM introduced the latest new function level, FL508, for Db2 12 for z/OS. This is the second new function level this year (the first came out in June and you can learn more about it here).

For those who don't know, the
Function Level process was designed by IBM for releasing new Db2 functionality using Continuous Delivery (CD) in short, quick bursts, instead of waiting for new versions (or releases). 

With FL508, IBM adds support for moving tables from multi-table table spaces, both simple and segmented, to partition-by-growth (PBG) universal table spaces (UTS). For an overview of UTS capabilities and types, check out this blog post I made earlier this year: Know Your Db2 Universal Table Spaces.

Multi-table table spaces are deprecated functionality, which means that even though they are still supported, they are on their way out. So it makes sense for IBM to give us a better way to convert them to PBG UTS without having to experience an outage. And that is just what FL508 delivers.

This is accomplished in FL508 by enhancements to the ALTER TABLESPACE statement. A new option, MOVE TABLE, is delivered which, as you might expect from its name, can be used to move a table from its current table space to a target table space. 

If, as you would expect in most cases, the source table space data sets are already created, the changes made by MOVE TABLE are pending changes and a REORG must be run on the source table space (the current one you are moving from) to materialize the change. Of course, this is an online REORG, so no outage is required.

The target table space must already exist as a PBG UTS in the same database as the current, source multi-table table space. Furthermore, the PBG UTS must be defined with MAXPARTITIONS 1, DEFINE NO, and [NOT] LOGGED and CCSID values that are the same as the current, existing table space. You can move only one table per ALTER TABLESPACE statement, meaning that each table in a multi-table table space must be moved with a separate ALTER TABLESPACE execution. However, because the changes are pending, you can issue multiple ALTER TABLESPACE statements, one for each table in the multi-table table space, and wait until they have all completed successfully before materializing all of the changes with a single REORG run. 

It seems simple, and the functionality is nice, but don't just go willy-nilly into things moving tables all over the place once you get this capability in FL508. IBM has documented the things to take care of before you begin to move tables using ALTER TABLESPACE. Check out the IBM recommendations here

It is also worth mentioning that you still need to keep in mind the impact that moving all tables from multi-table table spaces into their own table space will have on the system. By that I mean, you have to consider the potential impact on things like the number of open data sets (DSMAX ZPARM), DBD size, EDM pool size, and management issues (number of utility jobs, for example).

But it is nice that we now have a reasonable approach for moving tables out of deprecated multi-table table spaces so we can begin the process of moving them before they are no longer supported. A lot of shops "out there" have been waiting for something like this and it is likely to cause FL508 to be adopted quickly.

Let me know what you think by commenting below...

Wednesday, October 07, 2020

IDUG 2020 EMEA Db2 Tech Conference Goes Virtual

For those of you who have attended an IDUG conference before you know why I am always excited when IDUG-time rolls around again. And the EMEA event is right around the corner!

Participating at an IDUG conference always delivers a boatload of useful information on how to better use, tune, and develop applications for Db2 – both for z/OS and LUW. IDUG offers phenomenal educational opportunities delivered by IBM developers, vendor experts, users, and consultants from all over the world.

Unfortunately, due to the COVID-19 pandemic, in-person events are not happening this year... and maybe not for some time to come, either. But IDUG has gone virtual, and it is the next best thing to being there! The IDUG EMEA 2020 virtual event will take place November 16–19, 2020. So you have ample time to plan for, register, and attend this year.

If you attended any of the IDUG North American virtual conference earlier this year you know that you can still get great Db2 information online at an IDUG event. And there are a ton of great presentations at the EMEA virtual IDUG conference – just check out the great agenda for the event.

Of course, a virtual event does not offer the face-to-face camaraderie of an in-person event, but it still boasts a bevy of educational opportunities. And the cost is significantly less than a traditional IDUG conference: both in terms of the up-front cost (which is significantly less) and also because there are no travel costs... 

For just $199, you get full access to the virtual conference, as well as a year-long premium IDUG membership and a complimentary certification or proctored badging voucher. If you're already a premium member, you can add the EMEA 2020 Conference access to your membership for just $99.

You can register here

The Bottom Line

So whether you are a DBA, a developer, a programmer, an analyst, a data scientist, or anybody else who relies on and uses Db2, the IDUG EMEA Db2 Tech Conference will be the place to be this November 2020. 

With all of this great stuff available online from this IDUG virtual event, why wouldn’t you want to participate?

Thursday, September 17, 2020

Convert Your COBOL Db2 Programs to Java Without Rebinding

 As most Db2 developers and DBAs know, when you modify a Db2 program you have to prepare the program to enable it to be executed. This program preparation process requires running a series of code preprocessors that—when enacted in the proper sequence—creates an executable load module and a Db2 application package. The combination of the executable load module and the application package is required before any Db2 program can be run, whether batch or online.

But it is not our intent here to walk through and explain all of the steps and nuances involved in Db2 program preparation. Instead, we are taking a look at the impact of converting COBOL programs to Java programs, particularly when it comes to the need to bind as a part of the process.

We all know that issuing the BIND command causes Db2 to formulate access paths for SQL. If enough things (statistics, memory, buffers, etc.) have changed, then access paths can change whenever you BIND or REBIND. And this can be troublesome to manage.

But if the SQL does not change, then it is not technically necessary to bind to create a new package. You can prevent unnecessary BIND operations by comparing the new DBRM from the pre-compile with the previous version. Of course, there is no native capability in Db2 or the BIND command to compare the DBRM. That is why there are third-party tools on the market that can be used for this purpose.

But again, it is not the purpose of today’s post to discuss such tools. Instead, the topic is converting COBOL to Java. I have discussed this previously in the blog in the post Consider Cross-Compiling COBOL to Java to Reduce Costs, so you might want to take a moment to read through that post to acquaint yourself with the general topic.

Converting COBOL to Java and BIND

So, let’s consider a COBOL program with Db2 SQL statements in it. Most COBOL uses static SQL, meaning that the access paths are determined at bind time, not at execution time. If we convert that COBOL program to Java then we are not changing the SQL, just the code around it. Since the SQL does not change, then a bind should not be required, at least in theory, right?

Well, we first need to get into a quick discussion about types of Java programs. You can use either JDBC or SQLJ for accessing Db2 data from a Java program. With JDBC the program will use dynamic SQL whereas SQLJ will deliver static SQL. The Db2 BIND command can be issued using a DBRM (precompiler output) or a SQLJ customized profile. 

So, part of the equation to avoid binding is to utilize SQLJ for converted COBOL programs.

CloudFrame, the company and product discussed in the referenced blog post above can be used to convert COBOL programs into modular Java. And it uses SQLJ for the Db2 access. As such, with embedded SQLJ, static SQL will be used and the access paths will be determined at bind time instead of execution time.

But remember, we converted business logic, not SQL. The same SQL statements that were used in the COBOL program can be used in the converted Java. CloudFrame takes advantage of this and re-purposes the existing package from the previous COBOL program to the new Java SQLJ. CloudFrame automates the entire process as part of the conversion from COBOL to Java. This means that the static SQL from the COBOL program is converted and customized into SQLJ in java. This is a built-in capability of CloudFrame that allows you to simply reuse the same package information that was already generated and bound earlier.

This means no bind is required when you use CloudFrame to convert your Db2 COBOL applications to Java… and no access paths will change. And that is a good thing, right? Conversion and migration are already time-consuming processes; eliminating performance problems due to changing access paths means that one less issue to worry about during a COBOL to Java conversion when you use CloudFrame.

Wednesday, September 16, 2020

Planet Db2 is Back!

For those of you who were fans of the Planet Db2 blog aggregator, you’ll be happy to know that it is back up and operational, under new management.

For those who do not know what I am talking about, for years Leo Petrazickis curated and managed the Planet Db2 blog aggregator. Leo provided a great service to the Db2 community, but unfortunately, about a year ago he had to discontinue his participation in the site. So Planet Db2 has been gone for a while. But it is back now!

Before I continue, for those who don’t know what a blog aggregator is, it  a service that monitors and posts new blog content on a particular topic as it is published. This means that whenever any blog that is being tracked by the aggregator posts new content, it is highlighted with a link to the blog post on the aggregator site. The benefit is that you can watch the blog aggregator page (in this case Planet Db2) for new content instead of trying to monitor multiple blogs.

So if you are a Db2 DBA, programmer, user, vendor, or just an interested party, be sure to highlight and visit Planet Db2 on a regular basis to monitor what’s new in the Db2 blogosphere. And if you write a Db2 blog be sure to register your blog at the Planet Db2 site so your content is tracked and aggregated to Planet Db2… you’ll surely get more readers of your stuff if you do!


Tuesday, September 08, 2020

Know Your Db2 Universal Table Spaces

 One of the biggest changes in the last decade or so has been the introduction of new types of table spaces – known as Universal table spaces, or UTS. Not only are UTS new to Db2, they are quickly becoming the de facto standard type of table space for Db2 applications, new and old.

At some point, Universal table spaces will displace your existing segmented and classic partitioned table spaces. We’ll examine why this is so later in the post, but first let’s briefly describe what Universal table spaces are.

Two Types of Universal Table Spaces

Introduced in Db2 9 for z/OS, Universal table spaces combine the best attributes of partitioned and segmented table spaces. If you do not know what partitioned and segmented table spaces are, I refer you to this older article I wrote on DB2 Table Space Options to bring you up to speed (note that this article is almost 20 years old at this point).

Universal table spaces offer improved space management for variable length rows because they use space map pages (like segmented table spaces). Also, like segmented table spaces, UTS deliver improved mass delete performance, and you can immediately reuse the table segments after the mass delete. And like partitioned table spaces, Universal table spaces can grow large (up to 128TB of data) and consist of multiple partitions.

At a high-level, there are two types of Universal table spaces:

1.     Partition-by-growth (PBG): The PBG UTS creates new partitions as the amount of data grows without the need to specify key ranges. This type of UTS is beneficial for tables that grow over time and need the additional limits afforded by partitioning but can benefit from the performance of segmented.

2.     Partition-by-range (PBR): The range-partitioned, or PBR UTS requires a key range for partitioning like classic partitioned table spaces. A PBR UTS basically adds segmentation to the existing partitioned table space.


Both types of UTS can contain only a single table, but IBM presentations have indicated that this is likely to change at some point in the future (although nothing has been announced or confirmed for certain).

A partition-by-range UTS is basically a segmented, partitioned table space. The limit key ranges must be specified in the table DDL. Index partitioning, which was supported for the earliest classic partitioned table spaces, is not supported for a PBR UTS. So before converting your classic partitioned table spaces to PBR UTS, you must first convert from index-controlled partitioning to table-controlled partitioning. Check out this blog post for a trick to quickly convert to table-controlled partitioning.

The second type of UTS is the partition-by-growth Universal table space. As its name implies, a PBG UTS can automatically add a new partition as the data in the table space grows. Over time, as the UTS is used by applications, data gets added to the table. When the PBG UTS reaches its maximum size, a new partition is automatically added to the table space. The new partition uses the same characteristics as the existing partitions, including compression details, free space, and so on.

You control the type of UTS using the DDL keywords: NUMPARTS, MAXPARTITIONS, and SEGSIZE. To create a PBR UTS you specify both NUMPARTS and SEGSIZE. To get a PBG UTS you must code the MAXPARTITIONS and SEGSIZE parameters. MAXPARTITIONS indicates the limit on the total number of partitions that a PBG UTS can grow to. Be careful, because if you only code the NUMPARTS parameter without SEGSIZE, then you will create a traditional partitioned table space. If you only code the SEGSIZE parameter (without either NUMPARTS or MAXPARTITIONS) you will create a traditional segmented table space.

Db2 12 for z/OS

A significant new feature for supporting big data was introduced in Db2 12, relative page numbering (or RPN) for range-partitioned table spaces. An RPN range-partitioned table space can be created, or an existing range-partitioned table space can be changed to RPN via an ALTER TABLESPACE with PAGENUM RELATIVE, followed by an online REORG of the entire table space.

An RPN table space delivers many benefits for availability and storing large amounts of data. This requires an expanded RID, which increases from 5 bytes to 7 bytes.

From an availability perspective, you can specify DSSIZE at the partition level for RPN table spaces. Furthermore, the allowable DSSIZE value is no longer dependent on the page size and number of table space partitions. The DSSIZE change can be an immediate change (no online REORG required to take effect) as long as the change does not decrease the DSSIZE value. You still can decrease DSSIZE, but only at the table space level.

From a big data perspective, the DSSIZE can grow up to 1 TB for a partition. And the maximum table size increases to 4 PB with approximately 280 trillion rows per table. That is a lot of data that can be stored. Think about it this way: if you were to insert 1000 rows per second it would take more than 8000 years to fill the table to capacity!

Why Are Universal Table Spaces the Future of Db2?

As of today (September 2020, Db2 12 for z/OS), there are basically five types of table spaces from which to choose:

1.     Segmented table spaces

2.     Universal Partition-By-Growth (PBG) table spaces

3.     Universal Partition-By-Range (PBR) table spaces

4.     Universal Partition-By-Range Relative Page Number table spaces

5.     Classic partitioned table space

Of course, for new databases, it is best to remove the classic partitioned table space from consideration because the PBR UTS is more efficient (and classic partitioning will likely be deprecated at some point). Technically speaking, there are actually two other types of table spaces (LOB and XML table spaces), but they are not general-purpose table spaces and can be used only in specific situations (with LOB and XML data).

So why do I advise that you favor Universal table spaces over segmented whenever you can? Well, for a number of reasons. First of all, because Universal table spaces are newer and all you really need for most every Db2 implementation. Secondly, because many new features of Db2 can only be used with Universal table spaces. Newer features that only work with UTS include:

·     Clone tables

·     Hash-organized tables

·     Currently committed locking

·     Pending DDL

·     Inline LOBs

·     XML multi-versioning


And this trend is likely to continue. As IBM introduces new function levels and versions of Db2 with new features that only work with UTS, it will become increasingly difficult for DBAs to keep track of which table spaces are not UTS so that they can make sure they are not using any new features that will not work with their old types of table spaces.

What this means is that other than Universal table spaces, the only other type you should be using are segmented table spaces and then only when you absolutely must have a multi-table table space. Therefore, the best practice I recommend is to make all new table spaces Universal (except for multi-table table spaces which can be segmented).

So, what is the future of the segmented table space? For the immediate future, segmented table spaces will continue to be supported. My guess is that, at some point, IBM will deliver a multi-table UTS capability, and then at some point deprecate segmented table spaces. But this is only my guess. As of the date I am writing this, IBM has not committed to a multi-table UTS and the segmented table space is still the recommended (and only) method for assigning multiple tables into a single table space.

My general recommendation though is that you avoid multi-table table spaces unless you have many very small tables and are close to reaching the open data set limit (200,000). Of course, your limit may be lower depending on the setting of the DSMAX subsystem parameter, which specifies the maximum number data sets that can be open at one time. Acceptable values range from 1 to 200,000; a lower setting may be specified due to operating system contraints or storage/memory limitations.

My general recommendation for table spaces is to slowly work on a conversion project to migrate your classic partitioned table spaces to PBR UTS and your segmented table spaces to PBG UTS. Doing so will bring you to the latest and greatest Db2 table space technology and position you to be able to use all new functionality in current and future versions of Db2 whenever – and wherever – you see fit.


To make sure that your systems are up-to-date and ready for new functionality it makes sense to adopt Universal table spaces for all of your Db2 tables. The only exception is for multi-table segmented table spaces, and you shouldn’t have too many of them.

Good luck Universalizing your Db2 databases!

Wednesday, September 02, 2020

The Benefits of In-Memory Processing

One area that most organizations can benefit from is by better using system memory more effectively. This is so because accessing and manipulating data in memory is more efficient than doing so from disk.

Think about it… There are three aspects of computing that impact the performance and cost of applications: CPU usage, I/O, and concurrency. When the same amount of work is performed by the computer using fewer I/O operations, CPU savings occur and less hardware is needed to do the same work. A typical I/O operation (read/write) involves accessing or modifying data on disk systems; disks are mechanical and have latency – that is, it takes time to first locate the data and then read or write it.

There are many other factors involved in I/O processing that involve overhead and can increase costs, all depending upon the system and type of storage you are using. For example, each I/O consists of a multitude of background system processes, all of which contribute to the cost of an I/O operation (as highlighted in Figure 1 below). It is not my intent to define each of these processes but to highlight the in-depth nature of the processing that goes on behind-the-scenes that contributes to the costly nature of an I/O operation.

Figure 1. The Cost of an I/O

So, you can reduce the time it takes to process your mainframe workload by more effectively using memory. You can take advantage of things like increased parallelism for sorts and improve single-threaded performance of complex queries when you have more memory available to use. And for OLTP workloads, large memory provides substantial latency reduction, which leads to significant response time reductions and increased transaction rates.

The most efficient way to access data is, of course, in-memory access. Disk access is orders-of-magnitude less efficient than access data from memory. Memory access is usually measured in microseconds, whereas disk access is measured in milliseconds. (Note that 1 millisecond equals 1000 microseconds.)

The IBM z15 has layers of on-chip and on-board cache that can improve the performance of your application workloads. We can view memory usage on the mainframe as a pyramid, from the slowest to the fastest, as shown in Figure 2. As we go up the pyramid, performance improves; from the slowest techniques (like tape) to the fastest (core cache). But this diagram drives home our core point even further: that system memory is faster than disk and buffering techniques.

Figure 2. The Mainframe Memory Pyramid

So how can we make better use of memory to avoid disk processing and improve performance? Although there are several different ways to adopt in-memory processing for your applications, one of the best methods can be to utilize a product. One such product is the IBM Z Table Accelerator.

 IBM Z Table Accelerator is an in-memory table accelerator that can improve application performance and reduces operational cost by utilizing system memory. Using it can help your organization to focus development efforts more on revenue-generating business activity, and less on other less efficient methods of optimizing applications. It is ideal for organizations that need to squeeze every ounce of power from their mainframe systems to maximize performance and transaction throughput while minimizing system resource usage at the application level. You can use it to optimize the performance of all types of data, whether from flat files, VSAM, Db2, or even IMS.

 So how does it work? Well, typically a small percentage of your data is accessed and used a large percentage of the time. Think about it in terms of the 80/20 Rule (or the Pareto Principle).  About 80% of your data is accessed only 20% of the time, and 20% of your data is accessed 80% of the time.

The data that you are accessing most frequently is usually reference data that is used by multiple business transactions. By focusing on this data and optimizing it you can gain significant benefits. This is where the IBM Z Table Accelerator comes into play. By copying some of the most often accessed data into the accelerator, which uses high-performance in-memory tables, significant performance gains can be achieved. That said, it is only a small portion of the data that gets copied from the system of record (e.g. Db2, VSAM, etc.) into the accelerator.

High-performance in-memory technology products -- such as IBM Z Table Accelerator -- use system memory. Sometimes, if the data is small enough, it can make it into the L3-L4 cache. This can be hard to predict, but when it occurs things get even faster.

Every customer deployment is different, but using IBM Z Table Accelerator to optimize in-memory data access can provide a tremendous performance boost.

A Use Case: Tailor-Fit Pricing

Let’s pause for a moment and consider a possible use case for IBM Z Table Accelerator.

In 2019, IBM announced Tailored Fit Pricing (TFP), with the goal of simplifying mainframe software pricing and billing. IBM designed TFP as a more predictable, cloud-like pricing model than its traditional pricing based on a rolling-four-hour-average of usage. Without getting into all of the details, TFP eliminates tracking and charging based on monthly usage and instead charges a consistent monthly bill based on the previous year’s usage (plus growth).

It is important to note that last point: TFP is based on last year’s usage. So you can reduce your bill next year by reducing your usage this year, before you convert to TFP. Therefore, it makes a lot of sense to reduce your software bills to the lowest point possible the year before the move to TFP.


So what does this have to do with IBM Z Table Accelerator? Well, adopting techniques to access data in-memory can lower MSU usage – and therefore your monthly software bill. Using IBM Z Table Accelerator to optimize your reference data in-memory before moving to TFP can help you to lower your software bills and better prepare you for the transition to Tailored Fit Pricing.



 If you’d like to learn more about IBM Z Table Accelerator there is an upcoming SHARE webinar on September 15, 2020, that goes into some more details about the offering. It is titled Digital Transformation IncludesGetting The Most Out of Your Mainframe: click the link for details and to register to attend.






Friday, August 28, 2020

IBM Db2 on Cloud News

If you are thinking about, or have already adopted Db2 in the cloud, there is some recent news you should know about. But before we explore that news, let’s take a look at the quick highlights of using Db2 in the cloud.

IBM’s Db2 on Cloud offering is a fully-managed operational data store running the IBM Db2 11.5 engine offering 24x7x365 availability. So if you know Db2 on Linux, Unix, and Windows platforms, you know Db2 on Cloud… but there’s more.

Db2 on Cloud runs containerized Db2 on with a dedicated DevOps team managing the maintenance and updates required to run your mission-critical workloads. This includes features like seamless data federation, point-in-time recovery, HADR with multizone region support and independent scaling. So many of the administrative burdens of managing Db2 on-premises are handled by IBM in the cloud.

Now if you know me, and have been reading my “stuff” on cloud and DBA, you know that this does not mean that you can entirely offload you DBA. But it is cool and it does help, especially with DBA teams being stressed to their limits these days.

So yes, you can run Db2 on Cloud! And there are many good reasons to consider doing so, such as scalability, pay-as-you-use pricing, and to take advantage of managed services.

OK, So What is New?

I promised some news in the title of this blog post and so far we have just set the stage by examining IBM’s cloud offering of Db2 (albeit at a high level). So, what’s new?

Well, IBM is revamping its pricing plans. Before digging into the news, you need to know that IBM offers two high-level pricing plan options.

  • The Lite plan uses a shared multi-tenant system designed for application development and evaluation of IBM Db2 on Cloud. It is offered free-of-charge, without any time limitations.
  • Enterprise plans are for usage and deployment of business applications and systems. It includes one database per service instance with 4 vCPU x 16 GB RAM x 20 GB storage on dedicated compute slices with the option to use a three-HA-node configuration with multizone region support. Pricing starts at $989/month.

What is new is that on August 19 IBM introduced two new plans, the Enterprise non-HA plan, and the Standard non-HA plan.  This means that there are now four options, other than the free Lite plan: Enterprise HA, Enterprise non-HA, Standard HA, and Standard non-HA.

As is typical with IBM pricing, it is not really all that simple and it is getting more complex.  But options are always good (I think).

So what is this Standard plan that does not appear on the IBM Db2 on Cloud: Pricing page? Well, we can find this on the IBM Db2 on Cloud catalog page Here we see that (as one might expect) it is a lower-cost option between Lite and Enterprise starting at 8 GB RAM with 20 GB storage.

IBM also noted that IBM Db2 on Cloud is now available in the following six data centers: Dallas, Frankfurt, Tokyo, London, Sydney, and Washington. And your instances can be provisioned either with or without the Oracle compatibility feature.

It is important to note that IBM also notes that customers on older, legacy plans (how about that, cloud legacy already) will be required to upgrade their to one of the newer plans.


So, there are more options to choose from with your Db2 on Cloud implementations. And if you have an older plan take some time to familiarize yourself with the new pricing plan options and be ready to choose accordingly for your workload requirements.

Tuesday, August 18, 2020

Navigating the IBM COBOL 4.2 End of Service Waters: Chart a course to benefit your business


Surprisingly, COBOL has been in the news a lot recently, due to its significant usage in many federal govern-ment and state systems, most recently with unemployment systems, being in the news. With the global COVID-19 pandemic, those unemployment systems were stressed like never before with a 1600% increase in traffic (Government Computer News, May 12, 2020) as those impacted by the pandemic filed claims.

Nevertheless, there is another impending event that will likely pull COBOL back into the news as IBM withdraws older versions of the COBOL compiler from service. All IBM product versions go through a lifecycle that starts with GA (general availability), after some time moves to EOM (end of marketing) where IBM no longer sells that version, and ends with EOS (end of support) where IBM no longer supports that product or version. It is at this point that most customers will need to decide to stop using that product or upgrade to a newer version because IBM will no longer fix or support EOS products or versions.

Of course, code that was compiled using an unsupported COBOL compiler will continue to run, but it is not wise to use unsupported software for important, mission-critical software, such as is usually written using COBOL. And you need to be aware of interoperability issues if you rely on more than one version of the COBOL compiler.

So what is going on in the world of COBOL that will require your attention? First of all, earlier this year on April 30, 2020, IBM withdrew support for Enterprise COBOL 5.1 and 5.2. And Enterprise COBOL 4.2 will be withdrawn from service on April 30, 2022 – just about two years from now.

So now is the time for your organization to think about its migration strategy.

Why is COBOL still being used?

Sometimes people who do not work in a mainframe environment are surprised that COBOL is still being used. But it is, and it is not just a fringe language. COBOL is a language that was designed for business data processing, and it is extremely well-suited for that purpose. It provides features for manipulating data and printing reports that are common business requirements. COBOL was purposely designed for applications that perform transaction processing like payroll, banking, airline booking, etc. You put data in, process that data, and send results out.

COBOL was invented in 1959, so its history stretches back over 60 years; a lot of time for organizations to build complex applications to support their business. And IBM has delivered new capabilities and features over the years that enable organizations to keep up to date as they maintain their application portfolio.

So, COBOL is in wide use across many industries.

A majority of global financial transactions are processed using COBOL, including processing 85 percent of the world’s ATM swipes. According to Reuters, almost 3 trillion dollars in DAILY commerce flows through COBOL systems!

The reality is that more than 30 billion COBOL transactions run every day. And there are more than 220 billion lines of COBOL in use today. COBOL is not dead…

 What’s new in COBOL 6

With COBOL 5.1 and 5.2 already out of support, and COBOL 4.2 soon to follow, one migration path is to Enterprise COBOL 6, and IBM has already delivered three releases of it: 6.1, 6.2, and 6.3. There are some nice new features that are in the latest version(s) of IBM Enterprise COBOL, including:

  • Compile and runtime support delivering performance improvements for z15 hardware and z/OS 2.4 operating system
  • Increased compiler capacity making it possible to compile and optimize larger programs (6.1)
  • 64-bit (AMODE 64) support in this compiler enables users to process large data tables that require greater than 2 GB of addressing space (6.3)
  • JSON support (6.1) including JSON PARSE statement (6.2)
  • Support for many new features from the COBOL 2002/2014 programming standards including new statements like ALLOCATE, FREE and INITIALIZE; addition of Dynamic Length elementary items; conditional compilation using the DEFINE compiler option, and more
  • Many new compiler options
  • Improved usability with USS

At the same time, there are concerns that need to be considered if and when you migrate to version 6. One example is that the new compiler will take longer to compile programs than earlier versions – from 5 to 12 times longer depending on the optimization level. There are also additional work data sets required and additional memory considerations that need to be addressed to ensure the compiler works properly. As much as 20 times more memory may be needed to compile than with earlier versions of the compiler.

Some additional compatibility issues to keep in mind are that your executables are required to be stored in PDSE data sets and that COBOL 6 programs cannot call or be called by OS/VS COBOL programs.

And of course, one of the biggest issues when migrating from COBOL 4.2 to a new version of COBOL is the possibility of invalid data – even if you have not changed your data or your program (other than re-compiling in COBOL 6). This happens because the new code generator may optimize the code differently. That is to say, you can get different generated code sequences for the same COBOL source with COBOL 6 than with 4.2 and earlier versions of COBOL. While this can help minimize CPU usage (a good thing) it can cause invalid data to be processed differently, causing different behavior at runtime (a bad thing).

Whether you will experience invalid data processing issues depends on your specific data and how your programmers coded to access it. Some examples of processing that may cause invalid data issues include invalid data in numeric USAGE DISPLAY data items; parameter/argument size mismatches; using TRUNC() with binary data values having more digits than they are defined for in working storage; and data items that are used before they have been assigned a value.

Migration considerations

Keep in mind that migration will be a lengthy process for any medium-to-large organization, mostly due to testing application behavior after compilation, and comparing it to pre-compilation behavior. You need to develop a plan that best suits your organization’s requirements and work to implement it in the roughly 2-year timeframe before IBM Enterprise COBOL 4.2 goes out of support.

Things to consider:

  • Gartner research shows that “huge ‘all-or-nothing’ modernization programs often fail to meet expectations”
  • What is your current state? Which COBOL compilers are you using and what is your end goal (6.1, 6.2, 6.3)?
  • Remember that compiled programs will continue to run, so it may not be imperative to re-compile everything prior to the end of support date. Of course, it can be difficult to keep track of what has been converted and what has not if you do not have a plan moving forward other than “convert when the program has to be changed at some point.” And it can become difficult to keep track of all the requirements and incompatibilities for multiple versions of COBOL if you do not plan for, and eventually convert to a newer compiler version.
  • Do you have the COBOL talent and knowledge not only to convert but to continue supporting your existing portfolio of COBOL applications?
  • Enterprise application portfolios can be quite large, making it difficult to effectively discover and map all of the dependencies. Consider using tools to help. 

Migration challenges and Options to Consider

As you put your plan together, you might consider converting some of your COBOL applications to Java. An impending event such as the end of support for a compiler is a prime opportunity for doing so. But why might you want to convert your COBOL programs to Java?

Well, it can be difficult to obtain and keep skilled COBOL programmers. As COBOL coders age and retire, there are fewer and fewer programmers with the needed skills to manage and maintain all of the COBOL programs out there. At the same time, there are many skilled Java programmers available on the market, and universities are churning out more every year.

Additionally, Java code is portable, so if you ever want to move it to another platform it is much easier to do that with Java than with COBOL. Furthermore, it is easier to adopt cloud technologies and gain the benefits of elastic compute with Java programs.

Cost reduction can be another valid reason to consider converting from COBOL to Java. Java programs can be run on zIIP processors, which can reduce the cost of running your applications. A workload that runs on zIIPs is not subject to IBM (and most ISV) licensing charges... and, as every mainframe shop knows, the cost of software rises as capacity on the mainframe rises. But if capacity can be redirected to a zIIP processor, then software license charges do not accrue - at least for that workload.

Additional benefits of zIIPs include:

  • They are significantly cheaper to acquire than standard CPs
  • When workload is redirected to a zIIP it frees up capacity on the standard CP

So, there are many reasons to consider converting at least some of your COBOL programs to Java. Some may be worried about Java performance, but Java performance is similar to COBOL these days; in other words, most of the performance issues of the past have been resolved. Furthermore, there are many tools to help you develop, manage, and test your Java code, both on the mainframe and other platforms.

Keeping in mind the concerns about “all-or-nothing” conversions, most organizations will be working toward a mix of COBOL migrations and Java conversions, with a mix of COBOL and Java being the end results. As you plan for this be sure to analyze and select appropriate candidate programs and applications for conversion to Java. There are tools that can analyze program functionality to assist you in choosing which the best candidates. For example, you may want to avoid converting programs that frequently call other COBOL programs and programs that use pre-relational DBMS technologies (such as IDMS and IMS).

How to convert COBOL to Java

At this point, you may be thinking, “Sure, I can see the merit in converting some of my programs to Java, but how can I do that? I don’t have the time for my developers to re-create COBOL programs in Java going line-by-line!” Of course, you don’t!

This is where an automated tool comes in handy. The CloudFrame Migration Suite provides code conversion tools, automation, and DevOps integration to deliver very maintainable, object-oriented Java that can integrate with modern technology available within your open architecture.  It can be used to refactor COBOL source code to Java without changing data, schedulers, and other infrastructure components. It is fully automated and seamlessly integrates with the change management systems you already use on the mainframe.

The Java code generated by CloudFrame will operate the same as your COBOL and produce the same output. There are even options you can use to maintain the COBOL 4.2 treatment of data, thereby avoiding the invalid data issues that can occur when you migrate to COBOL 6. This can help to reduce project testing and remediation time.

It is also possible to use CloudFrame to refactor your COBOL programs to Java but keep maintaining the code in COBOL. Such an approach, as described in this blog post (Consider Cross-Compiling COBOL to Java to Reduce Costs), can allow you to keep using your COBOL programmers for maintenance but to gain the zIIP eligibility of Java when you run the code.

Upcoming Webinar

To learn more about COBOL migration, modernization considerations, and how CloudFrame can help you to achieve your modernization goals, be sure to attend CloudFrame’s upcoming webinar where I will be participating on a panel along with Venkat Pillay (CEO and founder of CloudFrame) and Dale Vecchio (industry analyst and former Gartner research VP). The webinar, titled Navigating the COBOL 4.2 End of Support(EOS) Waters: An expert panel discusses the best course of action to benefit your business will be held on September 23, 2020 at 11:00 AM Eastern time. Be sure to register and attend!


Users of IBM Enterprise COBOL 4.2 need to be aware of the imminent end of service date in April 2022 and make appropriate plans for migrating off of the older compiler.

This can be a great opportunity to consider what should remain COBOL and where the opportunities to modernize to Java are.  Learn how CloudFrame can help you navigate that journey.