CODEWITHSUNDEEP
databasecassandracassandra-2.0high-availabilitymulti-master-replication
nad2000
nad2000

Reputation: 4905

Solution for a small DB with high availability, no single point of failure

I need to choose a solution for a small, i.e., 2 tables/entities, in total max 100,000 rows (~ 20 MB of data) with HA having data distributed among multiple nodes. HA with no single point of failure is a must. I am considering Cassandra for the task. However, I feel that Cassandra might be an overkill for such a volume of data. Is there any lighter alternatives for this task that could provide HA with multi master replication capacity.

Upvotes: 2

Views: 454

Answers (1)

Andy Tolbert
Andy Tolbert

Reputation: 11638

I'd consider Cassandra a reasonable candidate for what you are trying to do. Replication and high availability is an out of the box feature for Cassandra where other database solutions tend to 'tack on' HA. Cassandra is also not too heavy duty and can run on small servers pretty dependably (see CassandraHardware)

On the other hand if you have not used Cassandra before you will have to do some learning to understand conceptually how replication, consistency and availability work in Cassandra, which may take some time if you are more familiar with these concepts in other database solutions.

With 20MB of data, it looks like Redis Sentinel could do the trick as well.

Upvotes: 3

Related Questions