Search Results hr_data_pump




Overview
Explore in AI Chat

Overview

APPS.HR_DATA_PUMP is a valid PL/SQL package body in the Oracle E-Business Suite Release 12.1.1 and 12.2.2 environments. It operates within the Oracle Human Resources (HR) and Oracle Payroll schemas, providing the underlying infrastructure for the HR Data Pump utility and the batch processing framework used by Oracle's HR legislative and payroll bulk data load operations. The package supplies a generic, metadata-driven batch loading engine into which partners and Oracle development teams plug their own validation and loading routines. Its principal design objective is to move large volumes of external data — typically supplied by third-party payroll bureaus, benefit providers, or statutory bodies — into HR and Payroll tables while preserving referential integrity, capturing line-level rejects, and reporting progress and errors to the request log. The ETRM record classifies the package as OTHER, indicating it is not an end-user API in the HR_API sense but an internal engine.

Key Procedures and Functions

The documented program units describe a complete envelope for batch execution:

  • ENTRY / EXIT — establish and release the package's runtime context, initialising globals and performing clean-up at the close of a run.
  • API_TRC_ON / API_TRC_OFF — toggle diagnostic tracing of the API calls issued during processing, used to debug partner-supplied load routines.
  • MESSAGE — writes informational, warning, or error text to the concurrent manager log via FND_FILE and registers messages through FND_MSG_PUB / FND_NEW_MESSAGES.
  • FAIL — raises the terminating condition for the current batch, recording the failure and assigning the appropriate status before transferring control back to the driver.
  • SLAVE — the worker entry point executed by each parallel child request; it consumes a range of batch lines and calls the registered load logic.
  • MAIN — the parent driver; it validates parameters, allocates work ranges, spawns slave requests, and consolidates results.
  • PURGEMAIN / PURGESLAVE — housekeeping routines that remove processed batch headers, lines, exceptions, and user keys for the parent and child processes respectively.

Tables Accessed

The package maintains its control information in the HR_PUMP_* tables. HR_PUMP_REQUESTS holds concurrent request registration; HR_PUMP_BATCH_HEADERS, HR_PUMP_BATCH_LINES, and HR_PUMP_BATCH_LINE_USER_KEYS store the staged input and its user-key mappings; HR_PUMP_RANGES partitions that input for parallel workers; and HR_PUMP_BATCH_EXCEPTIONS plus its _S sequence table capture rejected rows. It reads HR_API_MODULES to resolve which registered API or loader handles a given business process, FND_NEW_MESSAGES and FND_MSG_PUB for message text, FND_PROFILE for environment settings, and FND_REQUEST to submit slave jobs. HR_GENERAL, HR_UTILITY, HR_API, HR_MULTI_MESSAGE, HR_PUMP_UTILS, PAY_CORE_UTILS, PAY_CONTINUOUS_CALC, PAY_ACTION_PARAMETERS, and PER_BUSINESS_GROUPS_PERF provide supporting HR/Payroll services, while DBMS_SQL and PLITBLM handle dynamic SQL and index-by-table PL/SQL bindings.

Usage Notes

HR_DATA_PUMP is invoked indirectly rather than called from forms. Standard invocation is through the concurrent program that drives the HR Data Pump (parent and child phases) or through partner product code that registers a load routine in HR_API_MODULES and calls the pump's ENTRY/MAIN/EXIT envelope. The 267 objects that reference it — principally HR and Payroll legislative packages and country extensions — confirm its role as shared infrastructure. Customisations should not modify this package; instead developers register their own API modules and supply translate/validate/load logic that the pump orchestrates. Because the purge routines physically delete staged rows, installations should schedule PURGEMAIN/PURGESLAVE cautiously and retain adequate historical volumes of the batch tables.