Re: More fun with native RLA on an SQL view

James H. H. Lampert wrote:

Well, the query debug messages show nothing that strikes me as remarkable. The CPI432C/D listed a bunch of what I'm guessing are
SQL indexes (because the names end in the form, "I6," "I11," and
so forth), most of them with reason 17, one with reason 4, and
one ("...I15"), the one also mentioned in the CPI4328, with
reason 0.

There don't appear to be any messages associated with the record
that "slipped through."

James H. H. Lampert wrote:

<<SNIP>> for some reason, RLA is failing to honor the WHERE
clause in an SQL view (so far as I'm aware, it's only doing it
for one record in one file), and I'm at a complete loss to
understand why.

The query debug messages might give some insight; i.e. to
reproduce the scenario might require the same implementation.

Perhaps not so remarkable by itself. However if the problem occurs only when that index is being used by the query engine, in the manner described by that message [in its lower level details; i.e. CPI4328 for just its first level text, as with many messages, is not all that meaningful], _then_ that message might seem remarkable, or at least noteworthy :-)

<<SNIP>>

Regards, Chuck