You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When tables lack statistics, pseudo-objects are created for estimation
In scenarios with many tables (like 20 million tables), this leads to excessive memory usage
Each pseudo-object allocates two maps (for columns and indexes)
The issue affects more customers than just large cluster clients, including a user with only 100 tables. (It depends on the frequency of the queries that use the pseudo-objects)
Potential solutions include creating one global pseudo-object or modifying the estimation logic
Enhancement