Power Systems is the hardware platform on which the IBM i operating system runs. In addition to the foundational Power Systems hardware, IBM has also introduced the PureSystems concept – a converged infrastructure that combines functionality of an appliance with the flexibility of an arbitrarily scalable system.
IBM’s trademark relational database is integrated into IBM technologies and leverages the performance and virtualization capabilities of Power Systems hardware. DB2 for i is part of IBM’s family of DB2 products and supports a broad range of applications and development environments at a low cost of ownership due to its unique autonomic (self-managing) features. Contrast that with other enterprise databases that require frequent monitoring and re-allocation of resources to keep an infrastructure running properly.
Some features that make DB2 for i a solid choice for big data computing include:
Cost based query optimization that automatically selects the optimal access method for the query by calculating an implementation cost based on the current state of the tables referenced in the query and any access paths/indexes available.
Partitioned tables that allow more data in tables because each partition can have the maximum table sizes, exponentially expanding table sizes.
EVI (Encoded Vector Indexes) that simplify complex indexing environments by providing the performance advantages of a columnar database without affecting the underlying table
Other pertinent considerations that make IBM DB2 for i a “big data ready” database are LPG (Look ahead Predicate Generation), Spreading data automatically, Symmetric MultiProcessing and Omnifind
IBM i has a long history of being an extremely stable and reliable server-side operating system for hosting business applications on Power Systems hardware. The “i” stands for integrated and IBM i focuses on pre-integrated features put in place by IBM where most other platforms require separate third-party products to facilitate the same needs. Some features that are either unique to IBM i or where it is taken to greater extents compared to other operating systems include:
Single-level store rethinks computer storage as a single two-dimensional plane of addresses pointing to pages, which intentionally scatters the pages of all objects across all disks so objects can be stored and retrieved much more rapidly. As a result, an IBM i server rarely becomes disk bound.IBM i Chief Architect Steve Will talks single-level store
Single-level store rethinks computer storage as a single two-dimensional plane of addresses pointing to pages, which intentionally scatters the pages of all objects across all disks so objects can be stored and retrieved much more rapidly. As a result, an IBM i server rarely becomes disk bound.
Dynamic Virtual Server (f.k.a. Logical Partition) provides the capability to move processors, memory and interactive performance between virtual servers without requiring a restart.
Encryption of data on disk that protects data transmission to and from the disk drive, protects data transmission in the cross-site mirroring environment when the data being mirrored is on an encrypted independent disk pool and protects data in the case of theft of the disk drive. Backups can also be encrypted.
Simple-to-deploy, object-based security model
Virus-resistant object architecture
Subsystem Work Management that allows multiple workloads to run safely on one machine, lets applications share resources without colliding with one another and reduces management of multiple environments.Read more on subsystem work management from Steve Will, IBM i Chief Architect
Read more in two ITG reportsIBM i vs. x86 cost analysis