Spring data redis - multiple schema - change schema in session (getting the connection) VS. using multiple connection pools

Question

I find myself having to deal with more schema of a given Redis instance in the same application, using spring data redis 1.1.0.RELEASE and jedis client version 2.1.0.

At runtime, I have to use the right schema for querying the database, accordingly to a given rule (this is "given" and I cannot change it), which changes from time to time.

I'm wondering which of the following is the right approach:

1. in the session (redisTemplate.execute), try to retrieve the session's connection and change the DB index (the SELECT redis command, just for clarifying) just before beginning insertions;
2. keep multiple connection pools, one for each schema, then use the right pool instead of choosing the schema.

At a guess I feel the second as "the right way", but I would avoid to overload the application with too much pools. Which should I use? Have you got other insights?

Answer 1

I found some relevant information for you on this blog post:

Note: Though the database index is configurable, the JedisConnectionFactory only supports connecting to one Redis database at a time. Because Redis is single threaded, you are encouraged to set up multiple instances of Redis instead of using multiple databases within a single process.

This seems to indicate that it is best to have several redis connection factories each connecting to a different instance of redis.

P.S. I stumbled on you post because I had the same interrogations as you and the above blog post provides interesting information...