Adapter properties and commands

This topic describes the functions and usage of the Oracle Adapter commands and their options.

Adapter command summary

Use the adapter commands when specifying data sources and targets. The applicability of many of the commands depends upon whether you are specifying a source or target, whether a database/query file (.mdq) is used, and the situations in which the command applies.

Adapter commands can be used in GET > Source > Command or PUT > Target > Command settings in the Map Designer and Integration Flow Designer, using GET, PUT, DBLOOKUP, or DBQUERY function calls, or overriding a data source or target using execution commands in a RUN function or on the command line.

The following adapter commands are Oracle-specific database parameters.

Connect string

Specifies the host connect string. The corresponding adapter command is -CONNECT.

-CONNECT connect_string

connect_string: Specify the host connect string. If an SQL*Net host connection string is specified, the connection is established through SQL*Net. If none is specified, a direct connection is established to the database identified by the ORACLE_SID environment setting.

Database adapter type

This property specifies the database adapter type. The corresponding adapter command is -DBTYPE.

This command must be specified if the original card is not a database and no database is specified using the Database/Query adapter command (-MDQ) and the Database Name adapter command (-DBNAME).

-DBTYPE ORACLE

ORACLE: The database adapter type is Oracle.

Client result cache

Enables the Oracle adapter to retrieve the results of a SELECT query from the Oracle client result cache. The corresponding adapter command is -CACHE.

Trigger

Specifies the trigger string containing the options defined for a trigger specification. The corresponding adapter command is -TR (or -TRIG). This command is used to specify the trigger string for the trigger specification.

-TRIG trigger_string

trigger_string: Specify the trigger string containing the options defined for a trigger specification.

Row count

Specifies the number of rows to be retrieved per fetch. The corresponding adapter command is -ROWCNT.

-ROWCNT row_count

row_count: Specify the number of rows to be retrieved per fetch.

The time to fetch is optimized in accordance with the increase in the number of rows to retrieve. Therefore, a significant performance improvement may be gained by retrieving as many rows as possible in a single fetch.

The number of rows retrieved is limited only by the amount of memory installed on the querying computer.

The Oracle Adapter default for the number of rows to be retrieved per fetch is based upon an allocation of 64K buffer. A calculation is then performed to indicate the number of rows that can be held in the buffer.

-INSERTALL

Enables multiple rows of data to be inserted into an Oracle table at once. The corresponding adapter command is -IA (or -INSERTALL).

By batching a user-specified number of rows, execution times can be improved, when network latency exists between the Oracle adapter and the Oracle server. This option uses the Oracle INSERT ALL database statement, which allows multiple inserts with one database command, reducing network roundtrips during the output card processing of multiple input records.

The -INSERTALL command can be specified on output cards and works only on tables containing supported Oracle built-in data types. Custom types like Abstract Data Types (ADTs) will force the adapter to use one database insert operation at a time. This limitation applies when using the Default On NULL Oracle capability, as activated by the -DON command.

The -INSERTALL database statement constructed by the Oracle adapter can be viewed in the log file.

By default, a batch of ten inserts is used. However, this number can be overridden on the command line, depending on the network configuration and deployment requirements. The ideal number for performance will vary based on the network setup. The performance of various values can be compared using the audit option on the command line