Error Detection Optimization

[ah89]

The current error detection does not analyze denial constraints and calculates same tables for all of them we can divide them into two type symmetric and non-symmetric, also we can parallelize the queries (calculation in error detection is N^2 which N is the number of tuples)

Benefits: For the symmetric DCs, we reduce the calculation to half, also with parallelizing them we can increase speed even more

Costs: Two methods 1-determining the symmetric (which we already half of it implemented) 2- re-write the query function for symmetric function and one method to feed this queries in our parallelism framework