• The Four Hundred
  • Subscribe
  • Media Kit
  • Contributors
  • About Us
  • Contact
Menu
  • The Four Hundred
  • Subscribe
  • Media Kit
  • Contributors
  • About Us
  • Contact
  • Guru: How To Cancel A Bad SQL Update

    May 15, 2017 Ted Holt

    In Three Ways To Manage Unmatched Data I wrote about the use of the RAISE_ERROR function to force a SELECT statement to cancel when unmatched data is considered a fatal error. Another good use of RAISE_ERROR is to force an UPDATE statement to cancel when an invalid condition occurs.

    To illustrate, imagine that you and I work in a factory. All factories have inventory. The people we serve purchase some inventory items and manufacture others. Our job is to write a program that will allow certain people to zero out the inventory balance for certain types of purchased items.

    The users will enter a series of item numbers into a database table (physical file) named ItemBal3. Our program is to set the quantity on hand to zero, but only for type 3 and type 4 items. Our program contains this UPDATE statement:

    update items as i
       set i.QtyOnHand = 0
     where i.type in ('3','4')
       and i.itemnumber in (select item from ItemBal3);
    
    
    If the users enter the ID numbers of items that are of other types, the program ignores those items.

    Let’s take it a step further. Suppose that the presence of some other type of item in ItemBal3 is an error that cannot be overlooked. In such a case, we can make the UPDATE cancel itself, like this:

    update items as i                                            
       set i.QtyOnHand =                                                  
              case when i.type in ('3','4') then 0                      
                   else raise_error ('97905',                           
                             'Non-purchased items cannot be zeroed') end
     where i.itemnumber in (select item from ItemBal3)

    Notice that the item type is no longer tested in the WHERE clause, but in the SET. If the database manager attempts to modify a type-3 or type-4 item, the case expression returns zero, which is assigned to the QtyOnHand column.

    But if the database manager attempts to modify an item of another type, the system calls the RAISE_ERROR function, which cancels the UPDATE and returns SQL state 97905 to the caller. If the program is running under commitment control, the database manager rolls back any items that were changed before the invalid item was encountered. I can illustrate with an example.

    Here is the item master table:

    Item number Description Type Quantity on hand
    A-1 3-inch Doodle 1 20
    A-3 5cm Spinkler 2 20
    A-7 #7 Hoozit 3 20
    B-1 Widget, size 8 4 20
    B-2 #12 Skyhook 5 20
    F-3 4-inch Floozle 6 20

    Here is ITEMBAL3.

    Item
    A-7
    B-1
    F-3

    If the system updates the items in the order in which they are listed, the quantity on hand for A-7 and B-1 changes from 20 to zero. But when it tries to update F-3, RAISE_ERROR cancels the UPDATE statement. Under commitment control, the quantity on hand reverts to 20. Without commitment control, A-7 and B-1 have a zero balance after the canceled UPDATE.

    Under commitment control, database integrity is preserved. You can find the problem, fix it, and restart the program.

    Without commitment control, you don’t know what you have. Good luck.

    RELATED STORY

    Three Ways To Manage Unmatched Data

    Share this:

    • Reddit
    • Facebook
    • LinkedIn
    • Twitter
    • Email

    Tags: Tags: FHG, Four Hundred Guru, SQL

    Sponsored by
    Midrange Dynamics North America

    Want to deliver DevOps on IBM i?

    DevOps enables your IBM i development teams to shorten the software development lifecycle while delivering features, fixes, and frequent updates that are closely aligned with business objectives. Flexible configuration options within MDChange make it easy to adapt to new workflow strategies and policies as you adopt DevOps practices across your organization.

    Learn More.

    Share this:

    • Reddit
    • Facebook
    • LinkedIn
    • Twitter
    • Email

    COMMON Looking Youthful In 2017 HelpSystems Tackles IBM i Password Woes

    Leave a Reply Cancel reply

TFH Volume: 27 Issue: 33

This Issue Sponsored By

  • Profound Logic Software
  • Remain Software
  • ASNA
  • Linoma Software
  • Manta Technologies

Table of Contents

  • Open Source On IBM i: Let It Grow
  • HelpSystems Tackles IBM i Password Woes
  • Guru: How To Cancel A Bad SQL Update
  • COMMON Looking Youthful In 2017
  • The Five Things Clouds Need To Deliver For IBM i

Content archive

  • The Four Hundred
  • Four Hundred Stuff
  • Four Hundred Guru

Recent Posts

  • The Power11 Transistor Count Discrepancies Explained – Sort Of
  • Is Your IBM i HA/DR Actually Tested – Or Just Installed?
  • Big Blue Delivers IBM i Customer Requests In ACS Update
  • New DbToo SDK Hooks RPG And Db2 For i To External Services
  • IBM i PTF Guide, Volume 27, Number 33
  • Tool Aims To Streamline Git Integration For Old School IBM i Devs
  • IBM To Add Full System Replication And FlashCopy To PowerHA
  • Guru: Decoding Base64 ASCII
  • The Price Tweaking Continues For Power Systems
  • IBM i PTF Guide, Volume 27, Numbers 31 And 32

Subscribe

To get news from IT Jungle sent to your inbox every week, subscribe to our newsletter.

Pages

  • About Us
  • Contact
  • Contributors
  • Four Hundred Monitor
  • IBM i PTF Guide
  • Media Kit
  • Subscribe

Search

Copyright © 2025 IT Jungle