aidp-peoplesoft — Oracle PeopleSoft via AIDP aidataplatform

When to use

• User wants to ingest PeopleSoft data (HCM, FSCM, Campus Solutions, ELM, CRM) into a Spark DataFrame from an AIDP notebook.
• User mentions: "PeopleSoft", "PSFT", "PS HCM", "Campus Solutions".

When NOT to use

• For Oracle Autonomous DB family (ALH/ADW/ATP) → aidp-alh.
• For generic Oracle DB on Compute / Base DB / on-prem → aidp-oracle-db.
• For Oracle Siebel → aidp-siebel.

Prerequisites in the AIDP notebook

1. Helpers on sys.path (run aidp-connectors-bootstrap first).
2. Env vars / OCI Vault secrets:
   • PSFT_HOST (PeopleSoft DB host)
   • PSFT_PORT (typically 1521 for the underlying Oracle DB)
   • PSFT_DATABASE_NAME (Oracle SID / service)
   • PSFT_USER, PSFT_PASSWORD
   • PSFT_SCHEMA (typically SYSADM)
   • PSFT_TABLE (PeopleSoft record table, e.g. PS_JOB, PS_VOUCHER)

Read

import os
from oracle_ai_data_platform_connectors.aidataplatform import (
    AIDP_FORMAT, aidataplatform_options,
)

opts = aidataplatform_options(
    type="ORACLE_PEOPLESOFT",
    host=os.environ["PSFT_HOST"],
    port=int(os.environ["PSFT_PORT"]),
    database_name=os.environ["PSFT_DATABASE_NAME"],
    user=os.environ["PSFT_USER"],
    password=os.environ["PSFT_PASSWORD"],
    schema=os.environ.get("PSFT_SCHEMA", "SYSADM"),
    table=os.environ["PSFT_TABLE"],
)
df = spark.read.format(AIDP_FORMAT).options(**opts).load()
df.show(10)

Pushdown SQL

Push a complete source query at the database — connector skips schema/table option building and runs the SQL verbatim. Useful for joins, filters, derived columns where you don't want Spark to fetch the full table.

opts = aidataplatform_options(
    type="ORACLE_PEOPLESOFT",
    host=os.environ["PSFT_HOST"],
    port=int(os.environ["PSFT_PORT"]),
    database_name=os.environ["PSFT_DATABASE_NAME"],
    user=os.environ["PSFT_USER"],
    password=os.environ["PSFT_PASSWORD"],
    extra={
        "pushdown.sql": (
            "SELECT EMPLID, EFFDT, DEPTID, JOBCODE "
            "FROM SYSADM.PS_JOB "
            "WHERE EFF_STATUS = 'A' AND EFFDT >= DATE '2025-01-01'"
        ),
    },
)
df = spark.read.format(AIDP_FORMAT).options(**opts).load()
df.show(10)

Gotchas

• Connector is read-only. PeopleSoft is treated as a source-of-truth system; writes back must go through PeopleSoft's own application APIs, not Spark.
• Underlying Oracle DB. Most PeopleSoft installs run on Oracle DB; the connector hits the DB directly. Network reachability rules from aidp-oracle-db apply — cluster pod CIDR needs L3 path to the PeopleSoft DB host.
• SYSADM schema is the standard PeopleSoft owner. Ensure the connector user has SELECT privs on the PS_* tables you need — PeopleSoft tables are not granted to PUBLIC by default.
• Read row count before pulling — PeopleSoft fact tables can have hundreds of millions of rows. Always include a WHERE filter via pushdown.sql for production.

References

• Helper: scripts/oracle_ai_data_platform_connectors/aidataplatform.py
• Official sample: oracle-samples/oracle-aidp-samples → data-engineering/ingestion/Read_Only_Ingestion_Connectors/Oracle_PeopleSoft.ipynb