This shows you the differences between two versions of the page.
| Both sides previous revision Previous revision Next revision | Previous revision | ||
|
topics:grids [2009/11/13 03:50] els |
— (current) | ||
|---|---|---|---|
| Line 1: | Line 1: | ||
| - | ====== Omnidex Grids ====== | ||
| - | ===== Overview ===== | ||
| - | |||
| - | Omnidex Grids are a feature of the Omnidex product line that allow large databases to be distributed across multiple processors or servers. This improves performance by allowing a query to be processed on the fewest nodes possible using multiple concurrent processors. This approach allows great scalability and provides convenient mechanisms for adding new data to the database. | ||
| - | |||
| - | An Omnidex Grid consists of one or more computers used in combination to satisfy Omnidex queries. Some Omnidex Grid’s are placed on one or more large, multi-processor computers. Other Omnidex Grid’s are placed on an array of inexpensive, commodity, desktop PC’s. Both types of grids perform well, and each has distinct advantages. | ||
| - | |||
| - | An Omnidex Grid always will have a //grid controller// and a collection of //grid nodes//. The grid controller coordinates all activity on the grid and communicates between the user and the underlying //grid nodes//. The //grid nodes// perform all of the work of housing and indexing the data, and fulfilling the queries. The //grid controller// and the //grid nodes// may all be on the same computer, or they may be distributed between multiple //grid servers// as needed for performance and scalability. | ||
| - | |||
| - | Many Omnidex Grids are created as an Omnidex Snapshot, meaning that the underlying data is distributed into flat files across multiple directories or servers. These become very high-performing databases that can satisfy a wide variety of queries very quickly. Omnidex Grids can also be created against the underlying database, though the nodes then remain tethered to the servers hosting the database. | ||
| - | |||
| - | ===== Designing a Grid ===== | ||
| - | |||
| - | The primary considerations for designing an Omnidex Grid are the [[admin:grid:partitions | partitioning scheme]] and the [[admin:grid:distribution | distribution plan]]. The partitioning scheme determines how a large database will be split into different nodes. The distribution plan affects how the nodes will be distributed across multiple servers or multiple drives. Both of these design decisions have a substantial effect on performance. | ||
| - | |||
| - | ===== Creating a Grid ===== | ||
| - | |||
| - | Omnidex Grids are straightforward to create. Each grid node is a separate Omnidex environment, with an environment file, indexes and data. For convenience, there are methods to share a single Omnidex environment file across multiple nodes. There are also methods to share data files across multiple nodes. | ||
| - | |||
| - | The following steps must be completed to [[admin:grid:creation| create an Omnidex Grid]]. | ||
| - | |||
| - | ====== Quick Links ====== | ||
| - | {{page>:quicklinks&nofooter&noeditbtn}} | ||