Prepare
- Upload Oracle Essbase files
Guide
Essbase is a multi-dimensional (OLAP) database system designed for analytical and budgeting applications, storing data in cube format.
Planned Support
- Essbase versions 6.x through 11.x
- Dimension and hierarchy extraction
- Cube data normalization
- CSV export (flattened structure)
What You Get Out
Once the parser ships, DataMeans will extract your data into multiple modern formats:
| Output | Description |
|---|---|
csv/{TableName}.csv | One CSV file per table with all row data |
xlsx/{TableName}.xlsx | Excel workbook per table |
xls/{TableName}.xls | Legacy Excel format per table |
json/{TableName}.json | JSON array of records per table |
json/{TableName}.jsonl | Newline-delimited JSON (streaming-friendly) |
postgres.sql | PostgreSQL CREATE TABLE + INSERT statements |
schema/schema-graph.json | Relationship graph for visualization |
schema/er-model.json | ER model for diagram tools |
report.json | Structured extraction report |
report.md | Human-readable extraction summary |
File Requirements
.pag(page files containing cube data).ind(index files).db(database settings and metadata)essbase.cfg(server configuration)
Current Status
Parser development is in the planning phase. Multi-dimensional OLAP formats require specialized handling for dense/sparse block optimization.
Technical Notes
Essbase uses proprietary compression and block storage. Calculated members depend on runtime context and formula language that may not fully translate to relational exports.
Last updated: January 2026
Technical reference
Overview
Oracle Essbase is a multidimensional online analytical processing (OLAP) database that organizes business data in hierarchical cubes for rapid analytical queries and complex calculations. Originally developed by Arbor Software and acquired by Oracle through its purchase of Hyperion Solutions, it divides cube dimensions into dense and sparse groups so that data blocks are created only for member combinations that contain data. Unlike relational databases, Essbase stores data in cells within dimensional intersections, optimized for read-intensive analytical workloads like financial planning, budgeting, and scenario analysis.
History and Background
- 1992: Arbor Software releases Essbase, whose name derives from "Extended Spread Sheet dataBASE", pioneering commercial multidimensional databases.
- 1994: Arbor Software is granted a United States patent, filed in 1992, for storing and retrieving multidimensional data in computer memory.
- 1995: Arbor Software goes public on NASDAQ.
- 1998: Arbor Software merges with Hyperion Software, forming Hyperion Solutions.
- 2000: Essbase 6.0 adds attribute dimensions and direct I/O.
- 2004: Essbase 7.1 introduces the Aggregate Storage Option (ASO).
- 2005: Essbase is renamed Hyperion System 9 BI+ Analytic Services.
- 2005: IBM stops marketing DB2 OLAP Server, an OEM version of Essbase, which leaves support in January 2007.
- 2005: Information Age magazine names Essbase one of the ten most influential technology innovations of the previous ten years.
- 2007: Oracle acquires Hyperion Solutions for approximately $3.3 billion.
- 2008: Essbase 11.1.1 ships with Oracle EPM System 11, adding typed measures and varying attributes.
- 2014: Release 11.1.2.3.500 adds hybrid aggregation mode for block storage databases.
- 2017: Essbase becomes available as part of Oracle Analytics Cloud.
- 2019: Essbase 19c is released for deployment through the Oracle Cloud Infrastructure Marketplace.
- 2020: Essbase 21c restores downloadable installers for on-premises deployment.
- 2025: Release 21.8 ships in December as the current stable release.
File Format Specifications
Essbase uses proprietary binary formats for outline, data, and index storage.
File Extensions:
.otl- Database outline (dimensions and member properties).pag- Data pages (compressed data blocks, up to 2 GB per file).ind- Index files (pointers to data blocks).db- Database file (name, location, and database settings).esm- Essbase kernel file (block pointers and recovery information).tct- Transaction control file (one entry per transaction with its Active, Committed, or Aborted state).ddb- Partition definition file (saved in both the source and target database directories)
File Structure:
- Outline: Hierarchical dimension definitions
- Data Pages: Compressed blocks of cell data
- Index: One entry per stored sparse member combination, pointing to its data block
- Metadata: Calculation scripts (
.csc) and data load rules (.rul) - Numbering: Data and index files are named
essn.pagandessn.ind, where n runs from 1 to 65,535; a new file begins once one reaches the default 2 GB size limit - Compression: Data blocks use bitmap (default), RLE, or zlib compression; index value pair compression is applied automatically where appropriate
- Aggregate storage: ASO cubes store data in numbered
.datfiles within default, temp, log, and metadata tablespaces instead of.pagand.indfiles - Limits: Member, alias, and dimension names hold up to 1024 characters; application and cube names up to 30 characters
- Temporary: Restructure work files (
.otn,.pan,.inn,.esn,.tcu)
Key Components:
- Dimensions: Hierarchical structures (Time, Product, etc.)
- Members: Individual elements within dimensions
- Cells: Data values at dimension intersections
- Sparse: Dimensions with a low percentage of populated intersections
- Dense: Dimensions with a high probability of data at each intersection
Data Types and Structures
| Type | Description | Storage |
|---|---|---|
| DOUBLE | Numeric cell values | 8-byte double precision |
| TEXT | Text measures (release 11.1.1+) | Numeric ID mapped through a text list |
| TEXT LIST | Maps text measure strings to numeric IDs | Up to 32,766 values per list |
| DATE | Date measures (release 11.1.1+) | Numeric value, formatted on retrieval |
| BOOLEAN | Attribute dimension type | Two members (True/False) |
| NUMERIC | Attribute dimension type | Level 0 member names are the numeric values |
| MISSING | Cell with no data | #Missing marker |
Multidimensional Model:
- Cubes contain dimensions with hierarchies
- Members can have attributes and aliases
- Attribute dimensions are typed as text (the default), numeric, Boolean, or date
- Cells store calculated or input values
- Data blocks exist only for sparse member combinations holding data
- Dynamic Calc members are evaluated at query time rather than stored
Version Differences
| Version | Year | Key Changes | Compatibility |
|---|---|---|---|
| 1.0 | 1992 | Initial Arbor Software release | Client-server OLAP architecture |
| 6.0 | 2000 | Attribute dimensions, direct I/O | Block storage (BSO) engine only |
| 6.5 | 2002 | Hybrid Analysis maps relational sources | Mapped relational data not stored in cube files |
| 7.1 | 2004 | Aggregate Storage Option added | BSO and ASO storage engines |
| 9.0 | 2005 | Hyperion System 9 platform | Renamed BI+ Analytic Services |
| 11.1.1 | 2008 | Typed measures, varying attributes | Component of Oracle EPM System 11 |
| 12c | 2015 | Java agent replaces C agent | Bundled with Oracle BI 12c only |
| 19c | 2019 | REST API; web UI replaces EAS console | OCI Marketplace deployment |
| 21c | 2020 | On-premises installers restored | 11g applications migrate via LCM |
Compatibility Notes:
- Applications from release 11g migrate to 19c and 21c through LCM export and import utilities
- Hybrid aggregation mode lets block storage cubes use ASO-style dynamic aggregation
- Essbase 21c deploys on Oracle Cloud Infrastructure or on-premises servers
- Smart View replaced the classic Excel Spreadsheet Add-in, discontinued after release 11.1.2.2
- Essbase 19c and later supports Unicode-mode applications only
- Date attribute dimensions support dates from January 1, 1970, through January 1, 2038
Technical References
- Oracle Essbase Documentation
- Wikipedia: Essbase
- Wikipedia: Hyperion Solutions
- Database Administrator's Guide for Oracle Essbase
- Calculation and Query Reference for Oracle Essbase
To learn how to use this format with DataMeans, see the User Guide.