US20230084389A1 - System and method for providing bottom-up aggregation in a multidimensional database environment - Google Patents

Abstract

Systems and methods for supporting bottom-up aggregation in a multidimensional database computing environment. A dynamic flow, coupled with a data retrieval layer or data fetching component, which in some environments can incorporate a kernel-based data structure, referred to herein as an odometer retriever, or odometer, that manages pointers to data blocks, contains control information, or otherwise operates as an array of arrays of pointers to stored members, enables bottom-up aggregation of cube data which, for example with pure aggregating queries, provides considerable run time improvement.

Description

CLAIM OF PRIORITY AND CROSS-REFERENCE TO RELATED APPLICATIONS
• This application is a continuation of and claims the benefit of priority to U.S. patent application titled “SYSTEM AND METHOD FOR PROVIDING BOTTOM-UP AGGREGATION IN A MULTIDIMENSIONAL DATABASE ENVIRONMENT”, application Ser. No. 15/333,022, filed Oct. 24, 2016; which application claims the benefit of priority to U.S. Provisional Patent Application titled “SYSTEM AND METHOD FOR HYBRID SUPPORT FOR DYNAMIC UPPER LEVEL MEMBERS WITH A MULTIDIMENSIONAL DATABASE”, Application No. 62/245,892, filed Oct. 23, 2015; U.S. Provisional Patent Application titled “SYSTEM AND METHOD FOR HIGH PERFORMANCE BOTTOM-UP AGGREGATION IN A MULTIDIMENSIONAL DATABASE ENVIRONMENT”, Application No. 62/245,897, filed Oct. 23, 2015; U.S. Provisional Patent Application titled “SYSTEM AND METHOD FOR SUPPORT FOR LARGE QUERIES IN A MULTIDIMENSIONAL DATABASE ENVIRONMENT”, Application No. 62/245,901, filed Oct. 23, 2015; and U.S. Provisional Patent Application titled “SYSTEM AND METHOD FOR PROVIDING A MULTIDIMENSIONAL DATABASE”, Application No. 62/411,473, filed Oct. 21, 2016; and is related to U.S. patent application titled “SYSTEM AND METHOD FOR USE OF A DYNAMIC FLOW IN A MULTIDIMENSIONAL DATABASE ENVIRONMENT”, application Ser. No. 15/332,948, filed Oct. 24, 2016 and subsequently issued as U.S. Pat. No. 10,552,393 on Feb. 4, 2020; each of which above applications are herein incorporated by reference.
COPYRIGHT NOTICE
• A portion of the disclosure of this patent document contains material which is subject to copyright protection. The copyright owner has no objection to the facsimile reproduction by anyone of the patent document or the patent disclosure, as it appears in the Patent and Trademark Office patent file or records, but otherwise reserves all copyright rights whatsoever.
FIELD OF INVENTION
• Embodiments of the invention are generally related to multidimensional database computing environments, and are particularly related to a system and method for providing bottom-up aggregation in a multidimensional database environment.
BACKGROUND
• Multidimensional database computing environments enable companies to deliver critical business information to the right people when they need it, including the ability to leverage and integrate data from multiple existing data sources, and distribute filtered information to end-user communities in a format that best meets those users' needs. Users can interact with and explore data in real time, and along familiar business dimensions, enabling speed-of-thought analytics. These are some examples of the types of environment in which embodiments of the invention can be used.
SUMMARY
• In accordance with an embodiment, the system supports bottom-up aggregation in a multidimensional database (e.g., Essbase) computing environment. A dynamic flow, coupled with a data retrieval layer or data fetching component (which in some environments can incorporate a kernel-based odometer retriever, or odometer that manages pointers to data blocks, contains control information, or otherwise acts as an array of arrays of pointers to stored members) enables bottom-up aggregation of cube data which, for example with pure aggregating queries, provides considerable run time improvement.
BRIEF DESCRIPTION OF THE FIGURES
• FIG.1 illustrates an example of a multidimensional database environment, in accordance with an embodiment.
• FIG.2 illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• FIG.3 further illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• FIG.4 further illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• FIG.5 further illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• FIG.6 illustrates a process for use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• FIG.7 illustrates bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• FIG.8 further illustrates bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• FIG.9 further illustrates bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• FIG.10 illustrates a process for bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
DETAILED DESCRIPTION
• The foregoing, together with other features, will become apparent upon referring to the enclosed specification, claims, and drawings. Specific details are set forth in order to provide an understanding of various embodiments. However, it will be apparent that various embodiments may be practiced without these specific details. The enclosed specification and drawings are not intended to be restrictive.
• Multidimensional database environments, an example of which includes Oracle Essbase, can be used to integrate large amounts of data, in some instances from multiple data sources, and distribute filtered information to end-users, in a manner that addresses those users' particular requirements.
• FIG.1 illustrates an example of amultidimensional database environment100, in accordance with an embodiment.
• As illustrated inFIG.1, in accordance with an embodiment, a multidimensional database environment, operating as a database tier, can include one or more multidimensional database server system(s)102, each of which can include physical computer resources or components104 (e.g., microprocessor/CPU, physical memory, network components), anoperating system106, and one or more multidimensional database server(s)110 (e.g., Essbase Servers).
• In accordance with an embodiment, amiddle tier120 can include one or more service(s), such as, for example, provider services122 (e.g., Hyperion Provider Services), administration services124 (e.g., Essbase Administration Services), or studio/integration services126 (e.g., Essbase Studio/Essbase Integration Services). The middle tier can provide access, via ODBC/JDBC127,128, or other types of interfaces, to ametadata catalog129, and/or one or more data source(s)130 (for example, a relational database), for use with the multidimensional database environment.
• In accordance with an embodiment, the one or more data source(s) can also be accessed, via ODBC/JDBC132, or other types of interfaces, by the one or more multidimensional database server(s), for use in providing a multidimensional database.
• In accordance with an embodiment, aclient tier140 can include one or more multidimensional database client(s)142 (e.g., Essbase Server clients), that enable access to a multidimensional database (such as, for example, Smart View, Spreadsheet Add-in, Smart Search, Administration Services, MaxL, XMLA, CAPI or VB API Applications, Oracle Business Intelligence Enterprise Edition Plus, or other types of multidimensional database clients). The client tier can also include consoles, for use with services in the middle tier, such as for example anadministration services console144, or a studio/integration services console146.
• In accordance with an embodiment, communication between the client, middle, and database tiers can be provided by one or more of TCP/IP, HTTP, or other types of network communication protocols.
• In accordance with an embodiment, the multidimensional database server can integrate data from the one or more data source(s), to provide a multidimensional database, data structure, or cube(s)150, which can then be accessed to provide filtered information to end-users.
• Generally, each data value in a multidimensional database is stored in one cell of a cube; and a particular data value can be referenced by specifying its coordinates along dimensions of the cube. The intersection of a member from one dimension, with a member from each of one or more other dimensions, represents a data value.
• For example, as illustrated inFIG.1, which illustrates acube162 that might be used in a sales-oriented business application, when a query indicates “Sales”, the system can interpret this query as a slice or layer ofdata values164 within the database that contains all “Sales” data values, where “Sales” intersect with “Actual” and “Budget”. To refer to aspecific data value166 in a multidimensional database, the query can specify a member on each dimension, for example by specifying “Sales, Actual, January”. Slicing the database in different ways, provides different perspectives of the data; for example, a slice ofdata values168 for “February” examines all of those data values for which a time/year dimension is fixed for “February”.
Database Outline
• In accordance with an embodiment, development of a multidimensional database begins with the creation of a database outline, which defines structural relationships between members in the database; organizes data in the database; and defines consolidations and mathematical relationships. Within the hierarchical tree or data structure of the database outline, each dimension comprises one or more members, which in turn may comprise other members. The specification of a dimension instructs the system how to consolidate the values of its individual members. A consolidation is a group of members within a branch of the tree.
Dimensions and Members
• In accordance with an embodiment, a dimension represents the highest consolidation level in the database outline. Standard dimensions may be chosen to represent components of a business plan that relate to departmental functions (e.g., Time, Accounts, Product Line, Market, Division). Attribute dimensions, that are associated with standard dimensions, enable a user to group and analyze members of standard dimensions based on member attributes or characteristics. Members (e.g., Product A, Product B, Product C) are the individual components of a dimension.
Dimension and Member Relationships
• In accordance with an embodiment, a multidimensional database uses family (parents, children, siblings; descendants and ancestors); and hierarchical (generations and levels; roots and leaves) terms, to describe the roles and relationships of the members within a database outline.
• In accordance with an embodiment, a parent is a member that has a branch below it. For example, “Margin” may be a parent for “Sales”, and “Cost of Goods Sold” (COGS). A child is a member that has a parent above it. In the above example, “Sales” and “Cost of Goods Sold” are children of the parent “Margin”. Siblings are children of the same immediate parent, within the same generation.
• In accordance with an embodiment, descendants are members in branches below a parent. For example, “Profit”, “Inventory”, and “Ratios” may be descendants of Measures; in which case the children of “Profit”, “Inventory”, and “Ratios” are also descendants of Measures. Ancestors are members in branches above a member. In the above example, “Margin”, “Profit”, and Measures may be ancestors of “Sales”.
• In accordance with an embodiment, a root is the top member in a branch. For example, Measures may be the root for “Profit”, “Inventory”, and “Ratios”; and as such for the children of “Profit”, “Inventory”, and “Ratios”. Leaf (level 0) members have no children. For example, Opening “Inventory”, Additions, and Ending “Inventory” may be leaf members.
• In accordance with an embodiment, a generation refers to a consolidation level within a dimension. The root branch of the tree is considered to be “generation 1”, and generation numbers increase from the root toward a leaf member. Level refers to a branch within a dimension; and are numbered in reverse from the numerical ordering used for generations, with level numbers decreasing from a leaf member toward its root.
• In accordance with an embodiment, a user can assign a name to a generation or level, and use that name as a shorthand for all members in that generation or level.
Sparse and Dense Dimensions
• Data sets within a multidimensional database often share two characteristics: the data is not smoothly and uniformly distributed; and data does not exist for a majority of member combinations.
• In accordance with an embodiment, to address this, the system can recognize two types of standard dimensions: sparse dimensions and dense dimensions. A sparse dimension is one with a relatively low percentage of available data positions filled; while a dense dimension is one in which there is a relatively high probability that one or more cells is occupied in every combination of dimensions. Many multidimensional databases are inherently sparse, in that they lack data values for the majority of member combinations.
Data Blocks and the Index System
• In accordance with an embodiment, the multidimensional database uses data blocks and an index to store and access data. The system can create a multidimensional array or data block for each unique combination of sparse standard dimension members, wherein each data block represents the dense dimension members for its combination of sparse dimension members. An index is created for each data block, wherein the index represents the combinations of sparse standard dimension members, and includes an entry or pointer for each unique combination of sparse standard dimension members for which at least one data value exists.
• In accordance with an embodiment, when the multidimensional database server searches for a data value, it can use the pointers provided by the index, to locate the appropriate data block; and, within that data block, locate the cell containing the data value.
Administration Services
• In accordance with an embodiment, an administration service (e.g., Essbase Administration Services) provides a single-point-of-access that enables a user to design, develop, maintain, and manage servers, applications, and databases.
Studio
• In accordance with an embodiment, a studio (e.g., Essbase Studio) provides a wizard-driven user interface for performing tasks related to data modeling, cube designing, and analytic application construction.
Spreadsheet Add-in
• In accordance with an embodiment, a spreadsheet add-in integrates the multidimensional database with a spreadsheet, which provides support for enhanced commands such as Connect, Pivot, Drill-down, and Calculate.
Integration Services
• In accordance with an embodiment, an integration service (e.g., Essbase Integration Services), provides a metadata-driven environment for use in integrating between the data stored in a multidimensional database and data stored in relational databases.
Provider Services
• In accordance with an embodiment, a provider service (e.g., Hyperion Provider Services) operates as a data-source provider for Java API, Smart View, and XMLA clients.
Smart View
• In accordance with an embodiment, a smart view provides a common interface for, e.g., Hyperion Financial Management, Hyperion Planning, and Hyperion Enterprise Performance Management Workspace data.
Developer Products
• In accordance with an embodiment, developer products enable the rapid creation, management, and deployment of tailored enterprise analytic applications.
Lifecycle Management
• In accordance with an embodiment, a lifecycle management (e.g., Hyperion Enterprise Performance Management System Lifecycle Management) provides a means for enabling enterprise performance management products to migrate an application, repository, or individual artifacts across product environments.
OLAP
• In accordance with an embodiment, online analytical processing (OLAP) provides an environment that enables users to analyze enterprise data. For example, finance departments can use OLAP for applications such as budgeting, activity-based costing, financial performance analysis, and financial modeling, to provide “just-in-time” information.
Dynamic Flow
• In accordance with an embodiment, the system supports the use of a dynamic flow (referred to herein in some examples as Query Processing Dynamic Flow (QPDF)) in a multidimensional database (e.g., Essbase) computing environment. The dynamic flow process enables hybrid use of, for example, Aggregate Storage Option (ASO), Block Storage Option (BSO), or other types of storage containers, and provides a common flow to process a received input query in bottom-up mode. The approach can be used to reduce the size of the cube, which provides for efficient calculation of dynamic members.
• For example, in accordance with an embodiment, for queries accessing dynamic sparse members, the system can use an aggregate storage engine to satisfy the request. For queries that cannot be processed by the aggregate storage engine, the system can employ a block storage engine to satisfy the request, including for example bringing the data into an aggregate storage temporary tablespace.
• For example, in accordance with an embodiment, when executed by a computer system, the dynamic flow process can operate upon a multidimensional database to: (1) expand an input query to find all base/calculated data; (2) analyze the expanded query to find dependencies and an order of calculation; (3) define calculation units according to the preceding steps; (4) build a processing flow with the defined calculation units, and connect them; and (5) execute the processing flow, and determine a response to the input query.
• FIG.2 illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• In a typical multidimensional environment, in order to prepare the system to respond to input queries, the database server pre-calculates values for certain dimensions, and stores those pre-calculated values in the cube for later lookup.
• In accordance with an embodiment, when a dynamic flow is instead used, the ability to support dynamic query processing enables the database server to avoid pre-calculating and storing such values, which improves performance and reduces the storage of potentially empty cells.
• As illustrated inFIG.2, in accordance with an embodiment, the system can include one or more query processor(s)200, for example a Multidimensional Expressions (MDX)query processor202, and/or a SpreadSheet Extractor (SSE)204 query processor, that enablereceipt206 of aninput query208 from a client, to retrieve, access, or otherwise examine a set of data from a data source, as provided by and made accessible via the multidimensional database.
• In accordance with an embodiment, apreprocessor component210 can include adata retrieval layer212 or data fetching component (which in some environments can incorporate a kernel-based odometer retriever, or odometer or data structure stored in memory that manages pointers to data blocks, contains control information, or otherwise acts as an array of arrays of pointers to stored members), anaggregator component214, and acalculator component216, each of these layers and components can be provided as a software or program code that is executable by a computer system.
• Generally, described, in accordance with an embodiment, the preprocessor receives218 input queries, from the one or more query processor(s), for processing against the multidimensional database. The aggregator is adapted to perform hierarchical aggregations of the data. The calculator is adapted to perform calculations on the data, and to cooperate with the aggregator, as further described below, to utilize the data retrieval layer (including an odometer as appropriate) to at least one of populate and/or search within a cube, and to process a response to an input query.
• In accordance with an embodiment, the system can include one or more storage container(s)220, such as, for example, one or more of an Aggregate Storage Option (ASO)222, Block Storage Option (BSO)224, or other type ofstorage container226, each of which storage containers can act as an interface between the data that is read from/written to230 the data source or multidimensional database, and whichever data might be needed for aggregation and calculation at the preprocessor.
• FIG.3 further illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.3, in accordance with an embodiment, in response to the database server receiving an input query, the aggregator can operate240,242 in combination with the calculator, to process the query, as part of adynamic flow244, which can be similarly provided as software or program code that is executable by a computer system.
• For example, as illustrated inFIG.3, the dynamic flow process enables hybrid use of, in this example, one or more ASO, BSO, or other types of storage containers, and provides a common flow to process the query using these storage containers, in a bottom-up mode.
• In accordance with an embodiment, when the system begins to process an input query, it first determines, from an examination of the input query, which particular data or other information needs to be retrieved, i.e., a metadata. The system can then define246 for that input query, aninitial calculation unit250, which encapsulates an aggregation/calculation process that will retrieve a set of data from the storage container(s).
• In accordance with an embodiment, a data buffer260 (referred to herein, in some examples, as one or more output buckets) operates as a data structure into which each calculation unit can read/write data252, and which allows for a temporary storage of the data received from254 the storage container(s), for consumption by calculation unit(s).
• In accordance with an embodiment, when the dynamic flow is used with a BSO-type storage container, the dynamic flow process performs a pre-analysis of an input query, and an expansion of asked points to its base data.
• However, the amount of such expanded base data may be quite large.
• To address this, and to reduce the volume of expanded data, in accordance with an embodiment, a first dynamic aggregation can be performed during the fetching of data from the kernel, without a full expansion of the associated kernel structure (such as, for example, a kernel-side odometer as described above).
• In accordance with an embodiment, the dynamic flow then operates to expand the input query, to find all base/calculated data; and to analyze the expanded query to find dependencies and an order of calculation.
• FIG.4 further illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.4, in accordance with an embodiment, as each subsequent calculation unit262 is determined by the system as part of a dynamic flow instance, the aggregation/calculation process encapsulated by that calculation unit fills up263 another output bucket in the data buffer.
• In accordance with an embodiment, the dynamic flow process continues defining subsequent calculation units; and then connects the calculation units; and executes the resultant processing flow, to determine a response to the input query.
• FIG.5 further illustrates use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.5, in accordance with an embodiment, once the dynamic flow has been built, and has been processed to completion of a response, the preprocessor can create or otherwise populate266 adatabase cube270, and thedata response272 can be returned to the requesting client.
• In accordance with an embodiment, the dynamic flow can support a variety of modes: including streaming and non-streaming modes, as further described below.
• In accordance with an embodiment, in a streaming mode, each calculation unit includes two steps of execution, including that (a) each stored cell is processed according to a calculation unit logic (if it belongs to the unit) and the cell is returned to subsequent/next units and/or an output bucket, without a need for temporary storage; and (b) all calculated dynamic cells are then returned to next units and/or the output bucket.
• Advantages of the streaming mode include the provision of a streaming interface; and efficient data processing (for example, there is no need for temporary storage of unused cells, the mode requires low memory usage, and each cell will be processed only in relevant units).
• In accordance with an embodiment, in a non-streaming mode, each calculation unit stores all of its cells in a local temporary buffer, until all of the calculations are finished. Once all of the calculations are finished, then all calculated points are returned to the next unit, one-by-one (as in the streaming mode).
• Advantages of the non-streaming mode include a simpler logic for the calculation units and their interactions.
• FIG.6 illustrates a process for use of a dynamic flow with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.6, in accordance with an embodiment, atstep280, a multidimensional database environment is provided at a computer system, which enables data to be stored in one or more database cubes, and which enables queries to be received for data in the one or more cubes.
• Atstep282, a preprocessor, including an aggregator, calculator, and retrieval layer is provided, which operates according to a dynamic flow process, for querying one or more ASO, BSO, or other data storage engines, and which processes received input queries in a bottom-up mode.
• Atstep284, upon receipt of an input query, the input query is expanded, to find all base/calculated data.
• Atstep286, the expanded query is analyzed, to find dependencies and an order of calculation.
• Atstep287, calculation units are defined, according to the preceding steps.
• Atstep288, a processing flow is built, with the defined calculation units and connecting them in appropriate order.
• Atstep289, the processing flow is executed, and a response to the input query is determined.
Odometer Retrieval
• In accordance with an embodiment, the system supports bottom-up aggregation in a multidimensional database (e.g., Essbase) computing environment. A dynamic flow, coupled with a data retrieval layer or data fetching component (which in some environments can incorporate a kernel-based odometer retriever, or odometer that manages pointers to data blocks, contains control information, or otherwise acts as an array of arrays of pointers to stored members) enables bottom-up aggregation of cube data which, for example with pure aggregating queries, provides considerable run time improvement.
• In a multidimensional database environment, aggregation provides several advantages, together with some consequences. Although a kernel-based component, such as an odometer, can fetch stored and partitioned cells and return them to a next calculation unit; with larger sparse dimensions, larger data sizes, and queries for non-stored members, the system may need to pass a large amount of data from the kernel to, e.g., an aggregator.
• Dynamic members could be expanded, for example by producing rangesets of leaves' member numbers (memnos) for the dynamic members, followed by a merging of these rangesets into a single large rangeset, and providing that to the odometer retriever.
• However, for each stored cell which the odometer retriever passes to the aggregator, the recipient aggregator must generally search for that cell's ascendants among many rangesets of dynamic members (of which there may in turn be a number of ascendants), and perform aggregations to these ascendants.
• To address this, in accordance with an embodiment, a dynamic flow or process such as that described above, can be used to expand dynamic members down to a selection of aggregated members that are relatively easy to obtain.
• In this manner, the system can reduce the amount of data that must be passed from the odometer retriever.
• In accordance with an embodiment, the odometer can operate similar to an ASO-type flow, in that it need not output any cell, until all of the data is fetched.
• FIG.7 illustrates bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• In accordance with an embodiment, the odometer retriever can be used with a BSO-type storage container, in which a multidimensional cube may be represented by smaller building blocks; and wherein a particular block either has data, or else there is no data in that particular block.
• Since it would generally not be feasible to instantiate all possible blocks, a BSO-type storage container generally only creates those blocks that will contribute to the result.
• In accordance with an embodiment, the odometer enables the database server to process an input query in a dynamic fashion, including determining quickly which of the blocks in a cube contribute to that input query.
• For example, a query for cities in California may be processed to include only those California cities that will contribute values to the query.
• In accordance with an embodiment, an input query can be represented, e.g. in MDX, and can be used to define a region within the multidimensional database. This region may be illustrated as a subset of a very large cube, with links to other regions or cubes within the larger cube. The odometer provides a means of obtaining data from the larger cube, in a process referred to herein as odometer expansion. The database server can then retrieve the data provided by the odometer expansion, from block storage, to respond to the input query.
• As illustrated inFIG.7, when aninput query290 is processed, aninput odometer300 is created, and the system determines the blocks within storage for a retrieved part of theodometer304.
• This first step can be processed relatively quickly, since the database server can touch the data and determine whether it contributes to the result. If so, the system starts accumulating that data.
• FIG.8 further illustrates bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.8, odometer expansion is then used308, to create an expandedodometer310, including determining those blocks associated with the dynamic part(s) of theodometer312.
• Using the above approach, the odometer enables the system to scan block storage, retrieve appropriate blocks, and manipulate the numbers if necessary to produce the data asked for, which may not necessarily be the data scanned from storage.
• In particular, in accordance with an embodiment, the odometer enables the system to determine which blocks in a block-oriented storage contribute to the odometer, and scan those blocks while calculating an upper level, to further determine potentially contributing blocks, in a bottom-up manner.
• As illustrated inFIG.8, in accordance with an embodiment, the data indicated by theodometer expansion314 can then be determined within the cube, and provided as a response to the input query.
• FIG.9 further illustrates bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.9, an example odometer acts as an array of arrays of pointers to members. The dynamic flow operates upon an input odometer320, and expands the odometer for formulas, aggregations and solve-orders322. The input odometer can be sorted/grouped for units, keeping members that are needed for odometer retrieval first in dimensions, including, as illustrated in the example ofFIG.9, determining a retrieved part of theodometer324, and one or more dynamic parts of theodometer328,330.
• In the example odometer illustrated inFIG.9, the dimensions' order matches the outline, and each dimension is covered by at least a default/top member. The members for each dimension are sorted by memnoX numbers in ascending order. For BSO, the system can process sparse and dense dimensions' members that comply with one of the following: stored members; dynamic members with no formula, sum-only (“+” or “−”) aggregation; or members defined in a partition target that the dynamic flow does not expand.
• The example described inFIG.9 is provided for purposes of illustration. In accordance with various embodiments, other types of odometers and usages can be defined.
• FIG.10 illustrates a process for bottom-up aggregation with a multidimensional database, in accordance with an embodiment.
• As illustrated inFIG.10, in accordance with an embodiment, atstep340, a multidimensional database environment is provided at a computer system, which enables data to be stored in one or more database cubes, and which enables queries to be received for data in the one or more cubes.
• Atstep342, a preprocessor, including an aggregator, calculator, and retrieval layer is provided, which operates according to a dynamic flow process, for querying one or more ASO, BSO, or other data storage engines, and which processes received input queries in a bottom-up mode.
• Atstep344, an odometer that manages pointers to data blocks, contains control information, or otherwise acts as an array of arrays of pointers to stored members is provided, and an odometer retrieval is used to process an input query, including determining which blocks contribute to the query
• Atstep346, the odometer can touch or otherwise determine whether a data contributes to the result, and if so start accumulating that data while calculating upper level data.
Example Embodiment
• In accordance with an example embodiment, the system can include or operate according to various functionality or functions described below. In accordance with other embodiments, other types of functionality or functions can be included.
• In accordance with an example embodiment, the odometer is an array of arrays of pointers to members. The Dynamic Flow expands its input odometer for formulas, aggregations and solve-orders and then it's sorted/grouped for units, keeping members that are needed for Odometer Retriever first in dimensions.
• In accordance with an example embodiment, an ASO odometer (AD_QXASOMEMBER_T**) will be used for both ASO and BSO cases.
• 1. The dimensions' order matches the outline (as usual).
• 2. Each dimension is covered (as usual), by at least default/top (or dummy for attribute dimension) member.
• 3. The members for each dimension are sorted by their memnoX numbers in ascending order and are not duplicated.
• 4. For ASO, any member can present.
• 5. For BSO, the system shall process sparse and dense dimensions' members that comply one of the following:
• a. Stored members.
• b. Dynamic members with the following limitations on their sub-tree: no formulas, sum-only (“+” or “−”) aggregation, BSO solve-order must be 0.
• c. Members defined in transparent partition target that Dynamic Flow does not expand, i.e. these members shall be taken from partition regardless to other dimensions (e.g. partition on one dimension).
CoordType
• In accordance with an example embodiment, this is an offset type of a member inside its dimension component in an odometer. The array of these offsets determines the position of CMI inside the odometer.
CMI
• In accordance with an example embodiment, CMI is a current member [index] combination of a cell, or [rarely] a sub-cube (if some dimensions are ignored). CMI can be determined either by CoordType array, or by the actual pointers to the members pointed by the odometer.
Key Space
• In accordance with an example embodiment, a cell key is an integer scalar value [of specific width] corresponding to CMI's position inside the odometer. To provide both directions of this correspondence, this correspondence is a bijection. Typically, the forward transformation is a linear combination of vector of offsets inside the odometer with per-odometer vector of weights (weight per dimension component). The reverse is division with remainder by the vector of weights in specific order. The key width, the weights (and dimension order) are parts of specific Key Space instance. The Key Space instance translates CMI to a key and vice versa.
Main Module Flow (BSO)
• In accordance with an example embodiment, a main module flow (BSO) can include:
• 1. Intersect all transparent partition target regions with the odometer.
• 2. Obtain all remote connections.
• a. For each connection, there can be a number of remote odometers.
• 3. For each remote connection:
• a. Create Cells structures for cells, and create remapping structures. They can be either per remote connection or per remote odometer.
• b. Start remote transaction.
• c. For each remote odometer of that remote connection:
• i. Remap remote odometer's members according to transparent partition definition.
• ii. Fetch remote odometer cells to Cells structures.
• d. Close remote transaction.
• 4. Sequentially remap all received Cells structures (eliminating more than one remappings), like in Blocks Scan below, probably grouping cells “per block” for more effective remapping, and form resulting cells.
• a. If the odometer was expanded, remap to the original odometer coordinates.
• b. This handles dynamic calculations partially (as blocks may happen not to have local data).
Main Module Flow (ASO)
• In accordance with an example embodiment, for choosing the existing entry point function/class, at the first getNext( ) call, the system issues a wrapper function to fetch all needed cells into aso.buffer. The getNext( ) call gets next cell from aso.buffer. The key is decoded to coordinates. These coordinates are checked in isBelongToOutputBucket( ) against param.outputCoords for OUTPUT_COORD mask. If this cell should not go to param.outputBucket, it's immediately returned by the current getNext( ) call. Otherwise, the cell is put by putOutputBucket( ) to the param.outputBucket, and the system loops over another cell from aso.buffer until there are no more cells.
• The present invention may be conveniently implemented using one or more conventional general purpose or specialized computer, computing device, machine, or microprocessor, including one or more processors, memory and/or computer readable storage media programmed according to the teachings of the present disclosure. Appropriate software coding can readily be prepared by skilled programmers based on the teachings of the present disclosure, as will be apparent to those skilled in the software art.
• In some embodiments, the present invention includes a computer program product which is a non-transitory storage medium or computer readable medium (media) having instructions stored thereon/in which can be used to program a computer to perform any of the processes of the present invention. The storage medium can include, but is not limited to, any type of disk including floppy disks, optical discs, DVD, CD-ROMs, microdrive, and magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic or optical cards, nanosystems (including molecular memory ICs), or any type of media or device suitable for storing instructions and/or data.
• The foregoing description of the present invention has been provided for the purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations will be apparent to the practitioner skilled in the art.
• For example, while many of the embodiments described herein illustrate the use of an Oracle Essbase multidimensional database environment, in accordance with various embodiments the components, features, and methods described herein can be used with other types of online analytical processing or multidimensional database computing environments.
• The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, thereby enabling others skilled in the art to understand the invention for various embodiments and with various modifications that are suited to the particular use contemplated. It is intended that the scope of the invention be defined by the following claims and their equivalents.

Claims (15)

What is claimed is:

1. A system for providing bottom-up aggregation in a multidimensional database environment, comprising:

a data storage container including data blocks that represent a multidimensional database, for at least one of storage or analysis of data; and

a computing device having a processor executing instructions that cause the processor to:

receive an input query;

define, based on the received input query, a region of data within the multidimensional database which includes links to other regions within the multidimensional database;

create a data structure that includes pointers to the data blocks;

determine, by reference to the data structure, which of the data blocks contribute to processing the input query; and

scan the data blocks so determined, to calculate an upper-level of data blocks during processing of the input query.

2. The system ofclaim 1, wherein the storage container is one of a block storage option, or an aggregate storage option.

3. The system ofclaim 2, wherein a dynamic flow process includes a marking of upper-level members as dynamic, for use during the processing of the input query, and expanding dynamic members down to a selection of aggregated members.

4. The system ofclaim 3, wherein the data structure that includes pointers to the data blocks operates as an odometer that includes an array of arrays of pointers.

5. The system ofclaim 1, wherein the instructions cause the processor to provide a response to the input query, the response including the data blocks in the upper-level.

6. A method for providing bottom-up aggregation in a multidimensional database environment, comprising:

providing, at a computer system including a processor, a data storage container including data blocks that represent a multidimensional database, for at least one of storage or analysis of data;

receiving an input query;

defining, based on the received input query, a region of data within the multidimensional database which includes links to other regions within the multidimensional database;

creating a data structure that includes pointers to the data blocks;

determining, by reference to the data structure, which of the data blocks contribute to processing the input query; and

scanning the data blocks so determined, to calculate an upper-level of data blocks during processing of the input query.

7. The method ofclaim 6, wherein the storage container is one of a block storage option, or an aggregate storage option.

8. The method ofclaim 7, wherein a dynamic flow process includes a marking of upper-level members as dynamic, for use during the processing of the input query, and expanding dynamic members down to a selection of aggregated members.

9. The method ofclaim 8, wherein the data structure that includes pointers to the data blocks operates as an odometer that includes an array of arrays of pointers.

10. The method ofclaim 6, wherein the instructions cause the processor to provide a response to the input query, the response including the data blocks in the upper-level.

11. A non-transitory computer readable storage medium, including instructions stored thereon which when read and executed by one or more computers cause the one or more computers to perform the method comprising:

providing, at a computer system including a processor, a data storage container including data blocks that represent a multidimensional database, for at least one of storage or analysis of data;

receiving an input query;

defining, based on the received input query, a region of data within the multidimensional database which includes links to other regions within the multidimensional database;

creating a data structure that includes pointers to the data blocks;

determining, by reference to the data structure, which of the data blocks contribute to processing the input query; and

scanning the data blocks so determined, to calculate an upper-level of data blocks during processing of the input query.

12. The non-transitory computer readable storage medium ofclaim 11, wherein the storage container is one of a block storage option, or an aggregate storage option.

13. The non-transitory computer readable storage medium ofclaim 12, wherein a dynamic flow process includes a marking of upper-level members as dynamic, for use during the processing of the input query, and expanding dynamic members down to a selection of aggregated members.

14. The non-transitory computer readable storage medium ofclaim 13, wherein the data structure that includes pointers to the data blocks operates as an odometer that includes an array of arrays of pointers.

15. The non-transitory computer readable storage medium ofclaim 11, wherein the instructions cause the processor to provide a response to the input query, the response including the data blocks in the upper-level.

Images