We have a table with below PK
PRIMARY KEY ((customer_id),type)
we want to select all data related to specific type. select all and then filter by type in code is not efficient. If we have the ability to query specific node, then we can filter by clustering key.
is it a good idea to use allow filtering but combined with "PER PARTITION LIMIT 1"?
select [columns] from keyspace.customers where type='XXX' PER PARTITION LIMIT 1 allow filtering;