A few years ago,
I started keeping a little list of APARs that documented the “interesting” problems in DB2,
so that I did not have to go through hours of searching every month or two.
Over time it evolved into three distinct groups of corrections.
For any fixes relevant to Real-Time Statistics (As one of our prime products RealTimeDatabaseExpert relies on this data it behooves me to make sure it is as valid and good as can be!)
As the age old adage “garbage in = garbage out” is still correct when it comes to the DB2 Optimizer. Over the years, it has become very clever and at the same time very sensitive to dodgy data or correlations or, heaven forbid, manual well-minded but incorrect updates (To help here : we even wrote our own freeware tool Statistics HealthCheck to find the bad data – before it bites you!)
- SQL Performance group:
Here all of the tweaks and bugs of the Optimizer usually surface, but I have had to expand its remit over the years to now also include SQL Access Path and SQL Stats problems as these groups then cover all possible interesting things!
APAR list what’s new?
I have also changed how I store the data as it used to be in a PDS member for DB2 V6 and is now in three Excel spreadsheets for DB2 V8, 9 and 10. Soon I will drop V8 and add Vnext as V8 has now “gone” from the APAR point of view. I have also added useful columns like “last modified” so you can see if the APAR has moved since you last blinked, and the latest addition is a HIPER column. It used to be part of the description but could easily get lost and so now there is a Y for HIPER or Blank for normal.
So every month I trawl through the tech support database of DB2 z/OS and look for changed, updated, new, or closed APARs and then get all the data, pack it into my Excels and then mail it out to anyone who wants it. Now I know you can subscribe to lots of services that send out email alerts but I still find three Excels or comma separated files a *lot* easier to work with than an email and it is DB2 version self-documenting. No need to follow links to the bitter end of the IBM web site!
As a final thought on this topic, I also write some notes about which APARs I think are of real interest including HIPERs etc.
Get the three APAR lists:
So here is how the last set of data looked. The columns are always the same:
APAR, CLOSED date, STATUS, V8 PTF, 9 PTF, 10 PTF, HIPER, Description
The Status column can be blank (nothing to report), New, Closed, New & Closed (If IBM is quick!), Modified (If a closed APAR has been changed by IBM) or Fixed in Next (that you do not want to see!)
The PTF columns are either the PTF number, N/A for not applicable for this release or OPEN for, guess what?, open.
First is the RTS list:
Now the RUNSTATS list:
Here you can see one HIPER – and note that any typos come from IBM not me J(INDFINITELY…).
And finally the SQL Performance list:
If you are interested in getting these updates simply email me or firstname.lastname@example.org and ask!
As usual comments and queries are always welcome!