
|
Technical Fact Sheet Dharma SQLengine
Optimizations:
Transaction Processing Optimizations
- SMP (symmetric multi-processor) support provides transaction scalability.
- Compiled/cached SQL for repeated queries.
- Metadata and security information caching.
- Execution plan data pipelining.
- Pre-computed joins take advantage of built-in navigation elements.
Data Warehousing Optimizations
- Star-join schema optimization.
- Join order/method selection for complex decision support queries.
- Query rewrite techniques improve nested query performance.
- Join hints for influencing query execution plans.
- Multi-index/bitmap support for compound-predicate queries.
- Dynamic Indexing for multi-table joins.
- Index usage for scalar functions.
Statistics & Cost-Based Optimizations
- Table cardinality
- Query selectivity
- Query access costs
- Data distribution histograms
- Dynamic database statistics
Standards Conformance:
SQL-92
Complete SQL-92 entry level support and support for these SQL-92 Intermediate features:
Dynamic SQL
GET DIAGNOSTICS support
CASE expressions
INTERSECT, MINUS set operators
ALTER INDEX
Extensive character-string functions
SET SCHEMA
Date/time/interval types
CAST functionality
Variable-length character strings
Nested queries
Transaction isolation levels
Recursive views
Schema manipulation statements
SQL-92 Internationalization features:
Multi-byte characters
Multiple character-set support
Unicode support
Explicit collation sequences
Translation functions
Full SQL-92 data type support, including unstructured LONG types
Server-side Java Functionality:
Java Stored Procedures
Compiled and stored in the DBMS in bytecode format for direct execution without recompilation and parsing. Can import any JDK-supplied and third-party Java libraries. SQL statements are compiled and cached for optimum performance. Java stored procedures can return result sets to calling applications or procedures.
Java Triggers
Execute as a result of specified database modifications made through SQL or other interfaces.
Security
GRANT and REVOKE support: column, table, and view level
Inherits underlying DBMS-based user names
Inherits underlying DBMS-based security and authentication
ODBC 3.5
ODBC extended SQL grammar
ODBC Core, Level 1, and most Level 2 APIs
Extensive scalar functions
OLE DB
OLE DB v1.1 Driver Manager
JDBC 1.4
Fully supported, pure-Java driver
Universal Storage API
Index and navigation properties
- Mandatory components for multi-component indexes
- Supported sort orders
- Row-level locking support
- Stability of scan cursors
- Row-id sorting behavior
- Aggregate function capabilities
- Push-down predicate operator processing
- Multi-record append capability
- Multi-record fetch capability
- Supported relational operators
- Ability to update index
Statistics
Information about cost of access methods and distribution of data among and within tables
|