DCS is a NoSQL in-memory database product compatible with the Redis protocol. It features high performance, HA and scale-out, manages the memories of sub-databases and sub-tables, and supports disaster recovery, monitoring, and migration. DCS Redis greatly improves kernel performance, including but not limited to the following aspects:
l High reliability of active-standby nodes: A semi-sync mechanism is used to solve data loss caused by asynchronous synchronization in the native cluster.
l Cluster HA: self recovery of nodes, and automatic fault recovery. The standby node automatically takes over services in several seconds in case of a host crash.
l Optimistic locking mechanism: optimistic locking transactions and high-performance concurrent modification.
l Faster disaster recovery: no Gossip broadcast storms.
l Easy-to-use and self-developed Java clients: soft load balance, multiple load policies, simplified architecture complexity, and lower deployment costs.
l Self-developed snapshot + pipeline memory backup technology: less disk writes, controlled I/O traffic, smooth backup, and lower performance loss.
l Multi-dimension monitoring metrics: time consumption, error rate, and traffic.