HP Neoview Release 2.3 Software - Manual (Page 62)

• The set of targets (that you have set in the “Sources” (page 40) section) must be the same
for all sources in the job.
• If multiple job entries specify the same target, either all mapnames specified by the job entries
must be identical, or they must use the same targetformat, the same fieldmappings,
and a sourceformat with identical field names and field order. See
“Maps” (page 36) for
details about mappings.
• For options specified at the job entry level, settings (explicit or default) must be identical for
all job entries.
• For options specified at the “Sources” (page 40) level, input data sources must have identical
settings (explicit or default) for these options:
— File and Named Pipe Sources
◦ “forcestaging” (page 49)
◦ “sorted” (page 53)
— JDBC Sources
◦ “datasource” (page 47)
◦ “noaudit” (page 51)
◦ “password” (page 52)
◦ “system” (page 54)
◦ “url” (page 54)
◦ “user” (page 55)
• Database connections and transactions managed internally by Transporter are shared by all
job entries.
• Rows from different sources can be processed in the same transaction. There is no guarantee
that a specific number of rows from a given source is included in each transaction.
• For non-SQL dataformats referenced by maps to a specific SQL target, the fields must have
identical names and be in the same order.
Considerations for Extract Jobs
While extract jobs are specified in the control file and run with load jobs in the same control file
invocation, there are these differences from load jobs:
• Extract jobs cannot be recovered.
• You cannot use staging tables with extract jobs.
• Database isolation levels are different, depending on the source:
— For a table source, Transporter issues queries with an isolation level of READ
UNCOMMITTED.
— For a Neoview SQL query source, Transporter issues queries with an isolation level of
READ COMMITTED. However, you can change the isolation level by explicitly
specifying it in the query.
62 Load and Extract Operations