Oracle International Corporation

Adiel Yoaz

Qin Yu

Nipun Agarwal

Subramanian Muralidhar

Sam Idicula

Ravi Murthy

Bhaskar Ghosh

Sriram Krishnamurthy

Susan M Kotsovolos

Asha Tarachandani

Bhushan Khaladkar

Thierry Cruanes

Ashish Thusoo

Visar Nimani

Geeta Arora

Robert J Jenkins, Jr

Ajay Sethi

Shashaanka Agrawal

Techniques are provided to achieve performance improvements for path-based access to hierarchical data and for utilizing an extensible indexing framework. Extensible operators may be optimized by evaluating the operators using functions native to the database system instead of using functions defined through an extensible indexing mechanism. Furthermore, the database system may resolve a resource path of a query when the query is compiled, and may store the mapping of the resource path to one or more resources in a cursor for the query execution plan to avoid resolving the resource path each time the cursor is used to run the query. The cursor may be made dependent on the one or more resources to which the resource path refers at compile time. This takes advantage of the observation that the mapping of the resource path to particular resources rarely changes between evaluations of a particular query.

Techniques for efficient queries on a file system-like repository

An approach for using functional indexes as a primary filter is provided. A database query that contains a reference to a column of a table, but does not contain any reference to a functional index based on that column, is received by a database server. The database server adds to the database query a reference to the functional index on the column. The database server determines if the database query should be executed using the functional index. If the database server determines that the database query should be executed using the functional index, then the database query is executed with the functional index. If the database server determines that the database query should not be executed using the functional index, then data that indicates that the functional index is not to be evaluated when the database query is executed is stored.

Automatic use of a functional index as a primary filter

Techniques for partitioning nested tables are provided. A parent table includes a column for collection items. A nested table is created for storing items that belong to the collection items. The nested table is partitioned to create a plurality of nested table partitions. Each nested table partition is a distinct, separately stored structure within a database. How the nested table is partitioned may, or may not, be based on how the parent table is partitioned. For example, a nested table may be partitioned based on the same criteria in which the parent table is partitioned. As another example, a nested table may be partitioned, whereas the parent table is not partitioned.

Partitioning of nested tables

Dynamic return type generation in a database system. A request is received to execute a first function. A second function is executed if the first function is defined to return data in a first type of data structure. When executed, the second function returns formatting information that indicates an arrangement of fields of data within the first type of data structure. The first function is executed to obtain a collection of data formatted according to the first type of data structure, and the collection of data is organized according to the formatting information returned by the second function.

Dynamic return type generation in a database system

Techniques for executing flashback queries using domain indexes in a database system are provided. In one technique, a flashback parameter of the flashback query is not included in a call-out invocation of domain index routines that implement a domain index. A call-back query execution unit nonetheless retrieves, for call-back queries issued by the domain index routines, call-back query results that reflect the state of a database in the database system as of the time specified in the flashback parameter. In another technique, a call-out invocation of domain index routines includes the flashback parameter of a flashback query. The domain index routines issue call-back queries that require the retrieval of data which reflects the state of a database as of the time specified in the flashback parameter, which are executed by the call-back query execution unit.

Flashback support for domain index queries

A request to execute a query is received in a database system. In response, the database system determines that a collection of data elements to be returned by the query corresponds to a first data structure not defined within a type dictionary of the database system. The database system obtains attribute values that describe respective data fields within the first data structure, and records, within the type dictionary, a first data type definition that specifies the data fields described by the attribute values.

Query duration types

Techniques are provided for processing a database statement that includes a table function capable of parallel execution, where the input of the table function is specified in a subquery. During compilation of the database statement, a global cursor is generated for the database statement. The global cursor includes a first plan component for executing the table function and a second plan component for generating input for the table function. A 'dummy' cursor is then generated. When slaves execute the table function, they call the dummy cursor for input. When a slave calls the dummy cursor for input, the dummy cursor causes a context switch back to the global cursor, so that the second plan component within the first cursor generates the input for the slave.

Executing nested subqueries of parallel table functions in the parallel single cursor model

A method and mechanism for pipelined table functions is disclosed. With pipelining, data from a first table function does not have to be materialized into a complete collection before it is consumed by a succeeding table function. Instead, a producer of data creates a stream of data that is immediately utilized by a consumer of that data. Also disclosed is a method and mechanism for parallel processing of table functions, in which the set of work operated upon by a table function is sub-divided into smaller portions that are assigned to a plurality of table function slaves. Also disclosed is an integration between pipelining and parallelized execution for table functions.

Growing community of inventors

Foster City, CA, United States of America

Adiel Yoaz