Cassandra excels in a broad range of applications -- especially if you understand its data model and write your applications accordingly. It's an excellent choice for time-series data, and a poor choice for application queues. It performs the best if you can simply record history and compute from it, rather than going back and editing or deleting things a lot.
Its very well suited for storing graph types relationship information, such as a group of people and their relationships. Data modeling this sort of information in a traditional SQL database is a pain and inefficient. Using Neo4J allows for efficient modeling of data while providing rich querying capabilities using Cypher. Its also a great fit for any programming language because of its support for REST API. It's less appropriate for any other data structure other than Graph data. So as with any DB, evaluate the data structure and query and if the querying revolves around relationships, then Neo4J is a fit. If there is more need for looking up individual nodes and their associated information, Neo4J might not be the most efficient solution in the market.
High Availability - we utilize the data replication features of Cassandra. This enables us to access our data even when several nodes have gone down
Data Locality - our architecture combines Cassandra storage nodes and computation nodes in the same machine. This enables us to utilize data locality and limit expensive network IO to read data.
Elasticity - Cassandra is a shared nothing architecture. Nodes can be added very easily and they discover the network topology. As soon as a node has joined the Cassandra ring, the data is redistributed among the existing nodes and streamed to it automatically.
No Ad-Hoc Queries: Cassandra data storage layer is basically a key-value storage system. This means that you must "model" your data around the queries you want to surface, rather than around the structure of the data itself.
There are no aggregations queries available in Cassandra.
It would be nice to have some concept of namespaces, or some way of roughly making a single instance multi-tenant. It'd be nice to make sandboxing easier.
I would recommend Cassandra DB to those who know their use case very well, as well as know how they are going to store and retrieve data. If you need a guarantee in data storage and retrieval, and a DB that can be linearly grown by adding nodes across availability zones and regions, then this is the database you should choose.
Apache Cassandra has the best of both worlds, it is a Java based NoSQL, linearly scalable, best in class tunable performance across different workloads, fault tolerant, distributed, masterless, time series database. We have used both Apache HBase and MongoDB for some use cases which were within hadoop setup and JSON (JavaScript Object Notation) document store respectively, but given the overall factors favoring Apache Cassandra, it is a technology choice for multiple platforms!
I have not used anything like Neo4J because of how unique it is in the work that it allows me to do. I am not aware of any other graph database platforms and it might be because it is a growing area (especially in the world of pharmaceuticals). I would be open to trying other softwares though.
The open source version of Cassandra is only suggested for learning the basic concepts and play with its core features. Unless you really want to invest a lot in your developers and architects knowing every detail of Cassandra, I prefer the DataStax enterprise version. Although the license cost is relatively high, I think they it is worth it. I'm thinking about the support, the monitoring tool OpsCenter, and the integration of Solr and Spark (for data analysis).
Cassandra didn't fully replace our old and traditional relation database Oracle. In addition, it opens another door for us to deal with some special business use cases that NoSQL database can do better in a more feasible and efficient way.
For experimentation purposes, it had a positive impact on my company. It was very natural to work with Neo4j and so intuitive to visualize the data.
Neo4j community edition is free, which is what we experimented on. So there was no investment up front apart from employee's time. But this quickly gave results and it was time well spent.
Neo4j is a cool but very new technology. It was hard to have people onboard, especially some of the leadership and relational folks.
