Oracle TimesTen In-Memory Database is a memory-optimized relational database that empowers applications with the instant responsiveness and very high throughput required by today's real-time enterprises and industries such as telecom, capital markets, and defense. (Product Data Sheet)

Oracle TimesTen In-Memory Database delivers real-time performance by changing the assumptions around where data resides at runtime. By managing data in memory, and optimizing data structures and access algorithms accordingly, database operations execute with maximum efficiency, achieving dramatic gains in responsiveness and throughput, even compared to a fully cached disk-based RDBMS. The TimesTen in-memory database is designed as an embeddable database within the applications to further improve performance of database operations by eliminating inter-process communication and unnecessary network operations.

In-memory databases are often misperceived as limited to single-user applications and exposed to data loss when a server fails. Neither is true with Oracle TimesTen In-Memory Database. Multiple users and multi-threaded applications are common, using row-level locking with committed-read isolation. Oracle TimesTen databases are persistent and recoverable. Durability is achieved through a combination of transaction logging and database checkpointing to disk. Applications may choose the setting of their transaction ACID properties according to their performance requirements. Following the standard relational data model, SQL, JDBC and ODBC are used to access Oracle TimesTen databases. Any developer familiar with RDBMS and SQL interfaces will be immediately productive developing real-time applications with Oracle TimesTen In-Memory Database.

TimesTen Replication enables real-time transactional replication between TimesTen databases. Due to the mission-critical nature of the applications, most deployments add TimesTen Replication for high-availability and load balancing. Availability is an essential requirement for most real-time applications. Industries that operate 24x7, such as telecommunications, and global systems that are Web-accessible, such as travel and reservations sites, cannot tolerate service downtime. Securities trading systems must remain continuously available while financial markets are open.

TimesTen Replication follows a 'master-subscriber' model, whereby committed changes to designated tables or entire database are copied from their source to one or more subscriber databases. Flexible architecture enables different types of replication options to be deployed, including active-standby, active-standby pair with read-only subscribers, and active-active configurations. Replication is supported over LAN and WAN connections. In the event that bi-directional replication conflicts occur, a time-stamp-based collision detection and resolution mechanism prevents inconsistent replicas.

A transaction-log based replication scheme enables high efficiency and low overhead. Asynchronous and synchronous data replications are available. Asynchronous replication provides maximum performance, but the application is completely decoupled from the receipt process of the replicated elements on the subscriber. For 'pessimistic' applications that need higher levels of confidence that the replicated data is consistent between the master and subscriber databases, synchronous replication with the returnreceipt or return twosafe services are both supported.

For application-tier caching, the Oracle database option In-Memory Database Cache extends the TimesTen in-memory database to support updatable caches, managing data loading and synchronization between Oracle databases and Oracle TimesTen in-memory databases.

Related links

• Oracle In-Memory Database Cache Overview