Tag Info

About

DB2 is a family of database servers developed by IBM. The supported platforms for DB2 servers are z/OS (formerly OS/390), Linux, UNIX, and Windows (referred to collectively as DB2 for LUW), and IBM i (formerly OS/400 or i5/OS). Originally designed as a relational engine, IBM has expanded DB2's capabilities on some platforms to include native XML database features and an RDF-compliant graph store. On the IBM i platform, DB2 for i is tightly integrated into the operating system in a variety of ways, including a unique file system layer that allows database tables to be treated as structured files and vice versa.

DB2 for LUW has different editions: Express-C, Express Edition, Workgroup Server Edition, Enterprise Server Edition and Advanced Enterprise Server Edition. Each of these editions has different engine capabilities and different limits on the amount of RAM and the number of CPU cores that DB2 can use. DB2 for LUW offers a smooth upgrade path for databases that were created in a lower edition of the product (e.g., going from Workgroup Server Edition to Enterprise Server Edition).

DB2 Express-C is a no-cost DB2 engine that runs on Linux and Windows and is authorized for production use and redistribution. As part of the DB2 10.5 release in June 2013, IBM extended the system resource limits for DB2 Express-C to allow the database engine to use up to 16GB RAM and two CPU cores, and set a maximum size of 15 terabytes per database.

In October 2009, IBM introduced DB2 pureScale, a database cluster solution for AIX on System p, suitable for online transaction processing (OLTP) workloads. Linux x86_64 support for DB2 pureScale on IBM System x hardware was added in August 2010. IBM based the design of DB2 pureScale on the Parallel Sysplex implementation of DB2 data sharing on z/OS. DB2 pureScale provides a fault-tolerant, clustered database engine that uses low-latency RDMA messaging and GPFS shared storage to present the appearance of a single, monolithic database to the application layer.

To handle large data volumes and complex queries that are common to online analytical processing (OLAP), DB2 for LUW provides a shared-nothing clustering implementation known as the database partitioning feature (DPF), which uses hash key partitioning to evenly distribute a single copy of the database across multiple servers. Incoming queries are automatically multiplexed to access multiple partitions in parallel. DPF was introduced in 1994, as was at times rebranded as DB2 Parallel Edition, DB2 Extended Enterprise Edition, or InfoSphere Warehouse.

Free DB2 Books

history | show excerpt | excerpt history