Data migration and data validation

Data migration copies rows from a source system into Snowflake tables (historical loads, partitions, and optional incremental runs). Data validation compares what landed in Snowflake with the source (schema, metrics, and optional row-level checks). Both are run through the same Orchestrator and Workers model, driven by the Snowflake AIM Agent for Data Warehouses.

Note

Cloud data migration and validation always use Snowflake as the target. PostgreSQL (and other dialects) are supported as sources you migrate from, not as destinations you load into through these workflows.


Data migration and validation in the Snowflake AIM Agent for Data Warehouses

The agent’s guided workflow covers deployment, data movement, and checks after SQL conversion:

CapabilityWhat it means
Migrate dataCopy rows from source tables into Snowflake with automatic row-count validation as part of the flow
Validate dataCompare schema, metrics, and optionally rows between source and Snowflake after migration
migrate-objectsDeploys objects wave by wave; for tables, deploys to Snowflake then migrates data from the source

Procedure-level assurance uses baseline-capture and the migrate-objects test loop (two-sided validation of function and procedure output against captured baselines). That complements table-level migration and validation above.

Example prompt (from the Snowflake AIM Agent for Data Warehouses topic):

"Migrate data"
"Validate data"

For cloud table validation (schema, metrics, and optional row-level checks), ask the agent to run cloud data validation.


Supported source systems (cloud)

SourceCloud data migrationCloud data validationWorker connectivity highlights
SQL ServerYesYesODBC (or BCP when enabled)
Amazon RedshiftYesYesODBC; optional UNLOAD to S3
OracleYesYesOracle Instant Client + ODBC
PostgreSQLYesYesNpgsql (no ODBC); set ssl_mode in Worker TOML
TeradataYesYesteradatasql or ODBC; regular, write_nos, or tpt extraction for migration
Azure SynapsePlannedPlannedODBC (coming soon)

For the full capability matrix (code extraction, deploy, testing, and more), see Supported source systems in the Snowflake AIM Agent for Data Warehouses.

Note

Cloud workflows use the Snowflake Python connector (not the Snowflake CLI). Orchestrator and Worker hosts need Python 3.11+. Running validation consumes Snowflake credits for L2/L3 warehouse queries.