ScyllaDB supports OLTP use cases (so no need for rdbms)
It supports document storage (so no need for mongodb)
Columnar database (no need for cassandra).
I might be missing something here, but it seems that ScyllaDB is covering all the use cases for RDBMS, MongoDB and Cassandra and is faster then the NoSQL counterparts.
I still might use RDBMS for some use simple cases. But is ScyllaDB really a one shoe fit all DB?
While I think there is no “one shoe fits all” solution when it comes to databases in general (because each use case is slightly different and the needs are different) we still see lots of ScyllaDB users who replace multiple tools with just ScyllaDB. In this sense, ScyllaDB CAN be a one-shoe-fits-all solution if your goals are aligned with what ScyllaDB can offer.
ScyllaDB supports OLTP use cases (so no need for rdbms)
Yes it does and would be a great choice for any use cases where you require low latency, high availability, and large scale
It supports document storage (so no need for MongoDB)
You can store JSONs in ScyllaDB and you can query it. Your data model won’t be as flexible but ScyllaDB will provide better and consistent performance. We have a great user story on when to use ScyllaDB vs MongoDB
As far as comparing ScyllaDB to Cassandra, the interface is the same (CQL) but ScyllaDB is implemented with a different language (C++) and several benchmarks prove that ScyllaDB is more performant/less expensive than Cassandra.
I still might use RDBMS for some use simple cases
Yes I think that would make sense for a simple use case, and you can rely on ScyllaDB where performance and low latency matter.
Note that ScyllaDB is not a Columnar DB, nor is Apache Cassandra.
Both are considered “wide column store” (confusing term on its own).
ScyllaDB is not the best DB for a pure analytic DB (OLAP), no real-time, use case.
It is a good solution for mixing OLTP and OLAP, particularly with Workload Prioritization.