CODEWITHSUNDEEP
cassandraarchitecturedistributeddatabasenosql
Robin
Robin

Reputation: 1

Cassandra replication - replicate some data but keep some local

How do you configure Cassandra so that some tables are NOT replicated at all but others are? Is this actually a good use case for Cassandra?

I have a group of customers (max. 50) that will all supply data on a daily basis (~50,000 records per customer per day, ~200 fields per record). I need to pre-process the data to obfuscate sensitive information locally, then combine the data centrally for analysis and then allow reporting against the combined data set. I am planning on each customer having a local Cassandra node for the raw data load (several flat files), but I don't want this replicated until the obfuscation is complete. Can I do this with different tables spaces and replication factors? The data can be keyed using customer ID as a PK, if that helps.

Upvotes: 0

Views: 55

Answers (1)

mikea
mikea

Reputation: 6667

You could have a keyspace for the customer raw data with a replication factor of 1 and keep the raw data tables in there and then have the obfuscated data tables in a separate keyspace with a replication factor > 1.

Upvotes: 2

Related Questions