Currently we use Cassandra in production where we explicitly partition records. Around 30% of our queries are by partition. To give an analogy, assume a schema of Student(studentid, name, school, city). let’s assume PK is student id. and we partition by school and city. We have requirement to fetch records by School and City quite often. If I understand it correctly we need to create a secondary Index on School and City in Aerospike right? I wanted to know if Aerospike is suitable for this use case? Should we expect any performance issues for querying by secondary index?
Our major problem with Cassandra is it requires lots of server nodes and we have some classic anti patterns in place.
Any suggestions in this regard will be appreciated.