Well, in the current implementation, I'm specifically looking for a full v_found amount of records (or the end of the cursor) from each org unit. This makes the logic simpler to reason about, and for reasonable (read: likely and practical) limits (limit+offset) should not have a significant effect on performance.
Thinking about it now, I'm not liking my CONTINUE-based status and location checks. I think folding the location and status checks into the outer query is probably the better solution. Thoughts, before I push yet more commits?
Well, in the current implementation, I'm specifically looking for a full v_found amount of records (or the end of the cursor) from each org unit. This makes the logic simpler to reason about, and for reasonable (read: likely and practical) limits (limit+offset) should not have a significant effect on performance.
Thinking about it now, I'm not liking my CONTINUE-based status and location checks. I think folding the location and status checks into the outer query is probably the better solution. Thoughts, before I push yet more commits?