Create indexes necessary for uuid column type change

Why are we doing this work?

Create indexes necessary for uuid column type change asynchronously.

1. The end goal is to rename uuid_convert_string_to_uuid into uuid
2. To do that, we need to drop the uuid column.
3. This, in turn, will drop indexes present on the uuid column
4. If there's no unique index present on the uuid column our ingestion pipeline will break because it depends on an unique index - spec failure

Implementation plan

To mitigate this we will:

1. Create necessary indexes on uuid_convert_string_to_uuid asynchronously (this MR)
2. Create necessary indexes on uuid_convert_string_to_uuid synchronously (subsequent MR)