Thanks MT0 for the most helpful MATCH_RECOGNIZE hint. Sorry that I'm not so skilled with fiddle and for the extended delay getting back on this back-burner effort.

Regarding all advice about dates/times/strings, the source are tables TBTCO and TBTCP in SAP ABAP systems (nothing I can do there) - ours is 37TB+ and we wanted a way to mine job info for performance/etc. We have been collecting these 2 tables for a few years and due to the fact that JOBNAME and JOBCOUNT are not unique it is quite difficult to join them to see consolidated job step statistics.

In short, I'm quite sure this will be helpful to others so posting where I finally got: select <desired fields> from TBTCO a inner join TBTCP match_recognize( partition by JOBNAME, JOBCOUNT order by SDLDATE, SDLTIME, STEPCOUNT measures step_one.SDLDATE as STRTDATE, match_number() as match all rows per match pattern ( step_one next_step* ) define step_one as STEPCOUNT = 1, next_step as STEPCOUNT = prev(STEPCOUNT) + 1 ) b on ( a.JOBNAME = b.JOBNAME and a.JOBCOUNT = b.JOBCOUNT and a.STRTDATE = b.STRTDATE ) where a.STATUS = 'F'

Now we can have a consolidated historical view of ~250MM SAP job steps - which we will use in many helpful ways.