The DB2 Utility Conundrum
customer has been, or will soon be, faced with a decision to make on their DB2
utilities. As of DB2 Version 7, IBM decoupled the core utility functionality
from the base DB2 engine and is charging customers extra if they choose to use
IBM’s utilities. In all past releases of DB2, core utility functionality came
free with the DB2 base product. This is causing confusion and angst for many
mainframe DB2 customers.
all, let’s be clear on our terms. What is a utility versus other
popular terms such as “tool” and “solution.” A
is a single purpose batch program for moving and/or verifying database pages;
examples include LOAD, UNLOAD, REORG, CHECK, COPY, and RECOVER. A database
is a multi-functioned program designed to simplify database monitoring,
management, and/or administration tasks. A
is a synergistic group of tools and utilities designed to work together to
address a customer’s business issue.
Now, given those definitions, let’s examine the details of this DB2 utility
conundrum. IBM has invested a lot of time and development into the DB2 utilities
over the past few releases of DB2 – and IBM’s utilities work better and faster
than they used to. Why did this happen?
Independent software vendors like BMC Software, Computer Associates, and CDB
Software brought DB2 utilities to market that were easier to use, faster, and
delivered higher availability to DB2 applications. Actually, this is still true.
Driven by this competition – and the frustration of DB2 users – IBM was forced
to improve its utility functionality.
to recoup its investment, though, IBM could no longer “give away” free utilities
to its customers. So, now you have to buy them. Of course, currently IBM only
offers one choice to its customers: buy all of the DB2 utilities or none of
them. So, DB2 shops have to decide whether they will buy and run IBM’s
utilities, or use a competing set of utilities. The ISVs offer more flexibility,
though. For example, you could buy your REORG and UNLOAD from BMC, your CHECK
and COPY from Computer Associates, your LOAD from CDB, etc. It is not an
all-or-nothing proposition. But you will need to make sure you have the entire
suite of utility functionality covered. At a minimum you will need products:
to backup your data
to recover and
restore your backups
inefficient database objects
to collect database
statistics in the DB2 Catalog
to load and unload
to check for data
products may, or may not have the same or similar names as the IBM utilities
that perform the same tasks. Choosing the proper utilities with the required
functionality can be difficult though. The functionality of the IBM and ISV
utilities is not 100 percent exact. Many of the third party offerings still
provide more functionality at higher speeds than the IBM offerings. And in some
cases, the IBM utilities perform tasks that the ISV utilities do not.
troublesome is that many shops bought third party utilities from ISVs with the
comfort level of having the IBM utilities as a back up in case problems were
encountered. Additionally, most shops bought only a few of the third party
utilities, and not a full set. Now those shops may have to purchase additional
ISV utilities to get the same functionality they had before DB2 V7. Or they may
choose to buy both the IBM utility suite and the same third party utilities to
maintain a sort of status quo.
line is that DB2 customers have some decisions to make — do they buy IBM or ISV
utilities? Or perhaps both? With the end of service date looming for Version 6
(June 2005), the next six to nine months will see those in the DB2 community who
have not yet upgraded moving to Version 7 – and making the utility decision. It
is confusing. But each DBA staff must be able to understand the issues and
options available. Now is the time to examine the functionality, speed, and
ease-of-use of each set of utilities to make sure you are using the utilities
that best match your needs.
price will be a factor for most shops, too. But price should be examined as
total cost of ownership (TCO), not just initial cost of the utilities. Measuring
TCO can be difficult, but it should include things like CPU usage, time spent
creating utility jobs, ease of use features like restartability, job generation,
etc., as well as the initial cost.
And try to
keep a smile on your face even if it is difficult to understand why you have to
pay for something that you previously got for free.