The reason why we develop Pika is to solve redis's memory limitation
Pika used Nemo as storage engine, Nemo is developed base on rocksdb.
Nemo support multi data structure by using rocksdb's kv port. Such as list, hash, zset, set
Because rocksdb is using disk, so Nemo don't have capacity limitation.
The protocol used to contact with Pika is redis protocol, so you don't need to change your code to migrate to Pika
Pika used multi threads model. The main thread accept the request and then send the request to the work threads.