• August 17, 2022

Db2 Cloning Tool: What is Skeleton Cloning?

Unless you’ve seen a presentation on the Cloning Tool, you may not be aware of support for something called Skeleton Cloning. The only references to “skeleton” in the user guide are in the context of ISPF, and there’s no mention of it in the panels or profiles. 

Skeleton Cloning is a subset of Subsystem Cloning, where a portion of the subsystem is cloned (not the entire subsystem). This can be accomplished via volume-based cloning, or using the COPY-BY-DS command. There may be several use cases where Skeleton Cloning could be beneficial. One such use case is when migrating Db2 to a new release or catalog level (V12 & beyond). This use case is the primary story used to explain what Skeleton Cloning is and how to accomplish it. It can be found in a Redpaper that was published in November 2018 (REDP-5473). 

In this example, a clone of only the necessary components for migration, primarily the production Db2 catalog and directory, can be migrated away from the production environment to refine and prove your production migration processes and timing. Therefore, testing of the migration processes is done with a catalog and directory of the same size and pedigree of production, providing a higher degree of confidence that you are validating steps that will work in production. The example can be expanded to include application spaces, either via the Skeleton Cloning technique or with subsequent Application Cloning processes.

“DB2 Skeleton Cloning: Protecting Your Production Environment” is available for download here. At only 20 pages in length, it’s a quick read. Once you understand these concepts, you will have another cloning option at your disposal for the various requirements that come your way.

If this was helpful, please share it or comment below.

Happy Cloning!

Paul Bartak 7 Posts

Paul Bartak has been working with IBM databases since 1985, including the first version of DB2 on MVS. As an IBM customer for 14 years Paul worked in Application Development, Database Administration, and DBA. He joined IBM in 1999 as an Information Management Technical Specialist, teaming with the Sales force to present, design, install, demonstrate, and evaluate IBM Information Management solutions in a pre-sales environment; gaining multiple certifications in Database Administration, Data Replication, Blockchain, and Cloud Computing. Paul developed the four generations of the DB2 Migration Planning Workshops from DB2 V8 for z/OS, to DB2 11 for z/OS; and enablement for Db2 for z/OS pureXML and Temporal features. Currently, as a Distinguished Engineer with Rocket Software, Paul is focused on the tools used to maximize DB2 for z/OS value, with emphasis in database cloning and DevOps processes. He supports multiple Db2 User Groups, has been a speaker at several IDUG and Insight/Think conferences, is a Redbook/Redpaper author, and IBM Champion.


  • Nenad Vasiljevic Reply

    June 3, 2021 at 12:38 am

    Hello Paul,

    We have Cloning tool V3R2M0 and running on Db2 v12. Having problems comparing 2 application environments, that are using PeopleSoft with 10,000 tablespaces and 50,000 tables. Question is if this is possible or are there ways to achieve it? At the moment, tool works with smaller environments like 1,200 tablespaces but still can’t get it working with PeopleSoft environment.
    Hope to hear your suggestions soon.


    • V2 Reply

      August 19, 2021 at 8:33 am

      I apologize for this late reply. I was just made aware of your comment.

      It’s not clear from your comment as to what error you are receiving or if some other condition is being returned. Therefore, I’m replying with some generic comments which likely may apply to this type of situation.

      Application , or Table Space, Cloning does significant compatibility checking to ensure that the target will accept the cloned data from the source. As the number of source objects increases, this amount of work expands rapidly. This checking can uncover a number of different mismatch types. Many of these can be resolved by ensuring that REPAIR processing is enabled within the SET command of the Source job. Some require more involved resolution methods. I would need to know more to comment further.

      Speaking generally, it is possible to gain significant efficiency, and reduced validation checking, by using Subsystem Cloning over Table Space cloning when a large percentage of the subsystem/group is being cloned. Subsystem Cloning is generally the best choice for large ISV systems, like Peoplesoft, where the Db2 subsystem, or group, exists for that application. If shared with other Applications, then the benefits may erode a bit, but it’s probably worth a look. Subsystem Cloning can avoid many of the cross system checks because the Catalog & Directory are also being cloned. Therefore, problem objects, like simple table spaces, can be cloned with Subsystem Cloning; where they cannot be cloned with Table Space Cloning unless alternative copy methods are enabled.

      If you would like to discuss further, you can reach out to me directly at my email, pbartak@rocketsoftware.com.

      Paul Bartak

Leave a Comment

Your email address will not be published.