I currently have the following filters being used.
-- Do not touch anything in the mrt, raw, or bv schemas
SqlPackageFilter0=IgnoreSchema(mrt|raw|bv);
-- Do not touch anything in the "dl" schema with names that start with "vsa_" and "sf_"
SqlPackageFilter1=IgnoreName(dl,(vsa_|sf_).*);
With only these filters, it will attempt to drop existing child objects (3-part naming convention) such as the following:
[dl].[vsa_dialog].[ClusteredIndex_da42ede8a3924cddb04325ce5896413f]
(This will also occur for any statistics on the same table)
If I try to add a 3-part NameFilter such as IgnoreName(dl,(vsa_|sf_).*,.*)", it will work but it also does not allow for schemas to be created since the schema is only 1 part and it's evaluated from right to left which would match the last part .*. For example, when I want it to create a schema called "extstg" it will be ignored because it appears to match the last part of the IgnoreName filter.
The reason why the third expressions is .* is because the prefix for child objects can be anything they decide to name it.
How do we ensure that child objects such as statistics and indexes are ignored on tables that are already being ignored?
I currently have the following filters being used.
-- Do not touch anything in the mrt, raw, or bv schemas
SqlPackageFilter0=IgnoreSchema(mrt|raw|bv);-- Do not touch anything in the "dl" schema with names that start with "vsa_" and "sf_"
SqlPackageFilter1=IgnoreName(dl,(vsa_|sf_).*);With only these filters, it will attempt to drop existing child objects (3-part naming convention) such as the following:
[dl].[vsa_dialog].[ClusteredIndex_da42ede8a3924cddb04325ce5896413f]
(This will also occur for any statistics on the same table)
If I try to add a 3-part NameFilter such as
IgnoreName(dl,(vsa_|sf_).*,.*)", it will work but it also does not allow for schemas to be created since the schema is only 1 part and it's evaluated from right to left which would match the last part.*. For example, when I want it to create a schema called "extstg" it will be ignored because it appears to match the last part of the IgnoreName filter.The reason why the third expressions is
.*is because the prefix for child objects can be anything they decide to name it.How do we ensure that child objects such as statistics and indexes are ignored on tables that are already being ignored?