Is Cassandra for OLAP or OLTP or both?

Question

Cassandra does not match ACID like RDBMS, but CAP. So, Cassandra selects the AP from the CAP and leaves it to the user to configure the sequence. I definitely can't use Cassandra for transactions in the main transaction, because C * is a bit inconsistent. But Cassandra writes very quickly, which is good for OLTP. I can use C * for OLAP because reading is very fast, which is good for reports. So I realized that C * is only good when your application does not need your data to be consistent for some time, but should reading and writing be fast? If in my understanding it’s kind enough to list some applications?

Answer 1

ACIDs are properties of relational databases, where BASEs are properties of most nosql databases, and Cassandra is one of them. The CAP theorem simply explains the problem of consistency, accessibility, and tolerance of partitions in distributed systems. The good thing about Cassandra is that it has a custom consistency, so you can be pretty consistent (at the price of partition tolerance), so OLTP is doable. As phact said, there are even some banks that have built their transaction software on top of Kassandra. OLAP is also doable, but not only with Cassandra, as its partitioned row storage limits its capabilities. You must have something like Spark to fulfill complex queries.