HP Neoview Release 2.5 Software - Manual (Page 69)

The Transporter Job
The units that make up a control file are called jobs. A job is a load or extract operation consisting
of several job entries.
A job entry is a unique source/target pairing consisting of:
• A source for the data, named in the “Sources” (page 45) section of the control file
• A target for the data, also named in the “Sources” (page 45) section of the control file
• The mapping information you provide in the “Maps” (page 41) section of the control file
(for example, the table columns in which to put the data)
NOTE: If the control file encounters a duplicate source/target pair, Transporter exits with an
error.
Several jobs make up a Transporter control file. For more information about the control file, see
“Control File Organization and Syntax” (page 27) .
Each job is a load or an extract with your desired control file options applied. A control file can
have several load and extract jobs, but a job must be either load or extract. If a Transporter
operation fails, the job that failed is the recoverable entity. For more information about job
recovery, see
“Job Recovery” (page 80) .
Sources and Targets
Targets for load jobs and sources for extract jobs reside on the Neoview platform. These are the
possible source and target combinations for load and extract jobs:
Table 7-2 Sources and Targets for Loading and Extracting Data
Extract Load
Targets Sources Targets Sources
• Files
• Named pipes
• An existing SQL table on the
Neoview platform
• A Neoview SQL query
Data resulting from the
Neoview SQL query moves to
the designated target.
One of these, existing on the
Neoview platform:
• A Neoview SQL table
• An insertable view
• A synonym
• Files
• Named pipes
• JMS Trickle Feed
For information about Neoview SQL tables, views, and synonyms, refer to the Neoview SQL
Reference Manual.
Parallel Streams
The “parallelstreams” (page 62) option tells Transporter how many data streams to establish
between the client and the Neoview platform. This section discusses how the use of parallel
streams applies to load and extract jobs.
Parallel Streams for Load Jobs
Transporter manages the parallel streams and moves data to the appropriate columns in the
target table. Each stream is assigned to a subset of the table partitions.
Parallel Streams for Extract Jobs
If parallelstreams = 1, parallelism is disabled and Transporter creates one target file or
pipe.
To perform a parallel extract to multiple files or pipes:
The Transporter Job 69