I used a SPUFI to insert approx. 10,000 rows into a million row table. Then I used File Manager For z/OS (DB2) to eyeball my results. Everything looked good and then I slowly realized that my inserted records were always 21 rows apart as displayed.
I found that disconcerting because my belief is that the records come up sorted by key(s). So why would they be evenly spaced like that when I know the new keys are random? Looking closer I noticed that while all the old rows were coming up sorted on key(s) the new records I inserted were coming up as every 21st record within the old correctly sorted records. If I did an explicit sort on the keys then the new records were randomly spaced as expected.
So what is going on here, something analogous to blocking with flat files? I'm expecting a reorg overnight will eliminate this phenomenon.