OBI key terms

The core Oracle BI engine that parses logical SQL, builds and optimizes execution plans, queries underlying data sources, and applies aggregation and calculation rules before returning results to clients.

The binary metadata file that stores the three-layer model (physical, business, presentation) plus security and connection information used by the BI Server to translate logical requests into source queries.

The Microsoft Windows client used to create and edit the repository, including physical schemas, logical business models, presentation subject areas, variables, and security filters.

The repository layer that describes source tables, joins, keys, and connection pools, mirroring how data actually resides in each data source.

The repository layer that defines logical fact and dimension objects, measures, hierarchies, and the mappings that translate them into physical source columns and joins.

The repository layer exposed to users, organized as subject areas containing presentation columns and hierarchies that analysts drag into analyses and dashboards.

A mapping object that connects a logical table to one or more physical tables and specifies fragmentation content, aggregation levels, and column mappings for query generation.

A drill path defined in the repository that orders logical levels from general to detailed so users can navigate upward and downward through data in analyses.

A named position within a dimensional hierarchy, such as Year, Quarter, or Month, carrying column mappings and key definitions used for drill-down and aggregate matching.

The SQL dialect issued by BI clients such as Answers and dashboards, written against presentation columns and measures rather than physical tables, which the BI Server translates into source queries.

The source-specific query text the BI Server generates and sends to a database or other data source after optimizing a logical request.

A repository object holding connection details and credentials for a data source, allowing the BI Server to open, reuse, and limit concurrent sessions against that source.

A presentation-layer container grouping related columns and hierarchies for a business topic, giving users a scoped starting point when building analyses.

A user-facing field in a subject area that maps to a logical column or hierarchy level, hiding physical naming and structure from report authors.

An interactive report built in the web interface from criteria containing columns, filters, and views, which can be saved, shared, and used as the basis for dashboard content.

The query definition inside an analysis, listing selected columns, filters, sorting, and aggregation that the BI Server converts into a logical SQL request.

A presentation element within an analysis, such as a table, chart, pivot table, gauge, or narrative, that displays the result set in a chosen visual format.

A collection of pages containing analyses, prompts, links, and other objects arranged for at-a-glance monitoring; it can be personalized and shared with defined audiences.

An interactive filter control placed on a dashboard that passes chosen values to one or more embedded analyses, optionally constrained by a presentation column's value list.

A condition applied to an analysis or prompt that restricts returned rows, expressed against presentation columns and translated into source-level predicates by the BI Server.

The BI Server behavior of rewriting a query to use pre-aggregated summary tables when they satisfy the requested levels, reducing the volume of data scanned at run time.

A summary table holding precomputed results at a coarser grain, mapped in the repository so the BI Server can substitute it for detailed queries when levels match.

A shared memory and disk store of previously executed logical query results that the BI Server reuses when an identical request and unchanged data make recomputation unnecessary.

The action of invalidating cached results, done manually or triggered by detected source changes, so subsequent queries are recomputed rather than served stale data.

A repository-configured table the BI Server polls periodically to detect source data changes and automatically purge affected cache entries.

A repository object that runs a query at session start or on demand to populate session or repository variables used in security filters and dynamic logic.

A variable whose value is set per user session, commonly by an initialization block, and referenced in filters or calculations to personalize results.

A variable with a single value shared across all sessions, used for constants and dynamic references that must stay consistent for every user.

A named set of permissions in the web catalog that controls access to dashboards, analyses, and folder objects for assigned users or groups.

A repository-level security expression attached to a user or group that silently restricts which rows a user can see in every query against a subject area.

The file-based store of web objects such as analyses, dashboards, prompts, and folders, with permissions managed through roles and object-level access settings.

A validation routine in the Administration Tool that reports repository errors and warnings, such as unmapped columns or invalid joins, before the repository is deployed.

A connection pool setting that lets many concurrent BI Server requests share a smaller number of physical database connections, reducing source-side load.

The internal representation of how the BI Server will satisfy a logical request, showing source queries, joins, aggregation steps, and any cache or aggregate substitutions.