問題文
An insurance carrier keeps millions of claim lines as children of each policy record. Dashboards that total premium and count open claims per policy take too long, so the architect proposes storing the total and the count on the policy record itself and maintaining them with automation on every child change. Which statement best describes the cost the team accepts by doing this?
選択肢
- Nothing is given up, because stored totals are recalculated by the platform whenever a report reads them, so the automation exists purely to give the dashboards a column that they are able to sort on.
- The only real cost is storage, because two numeric fields on the parent consume space while the totals themselves stay accurate without any further work once the automation has been created, and the dashboards read the two fields directly.
- Denormalizing removes the need for the relationship itself, so the child object can be retired once the totals have been stored on the parent, and the detail is only needed again if the business later decides to audit how each stored figure was produced.
- Reads get faster, but the org now owns the correctness of the stored totals: every child insert, update, and delete has to write the same parent row, which adds contention and leaves the totals wrong whenever one update path is missed.