OBIEE - Query Miss the Cache Hit

Yes, a superset logical query that is seeded can satisfy a number of other queries that return the subset of the superset. But there are a lot of limitations and reason for missing the hit on an already seeded superset cache and are given below;

1.    Cache hit will miss if the aggregation rules are applied at report level rather than implementing at repository (RPD) level.

e.g. AVG, AvgDistinct, BottomN, Count, CountDistinct, Count (*) , First, Last, Max, Median, Min, NTile, Percentile, PeriodAgo, PeriodToDate, Rank, StdDev, Sum, SumDistinct, TopN etc.

By default these aggregate functions are grouped at the lowest level however can be grouped on some other levels using the BY statement in the function definition.

2.    The logical query/report has an additional column as compared to the seeded report/query

3.    The logical query/report has a where clause on the column that is not available in the seeded query/report

4.    There are very limited aggregation possibilities available e.g. The logical query/report contains a column plan_payment and is cached, a request for sum(plan_payment) will miss the cache hit.

5.    The relationship operator in the filters does not qualify the subset of a superset.

6.    The logical query/report is created with a distinct superset and cached. The cache hit will miss if the logical query/report eliminates the distinct processing.

As a matter of fact the cache seeding and hitting in case of subset of a superset works in relatively simpler cases.