Pushdown Processing
What is pushdown processing
Pushdown processing delegates profiling and DQ evaluation to your connected data source instead of running them on Ataccama’s runtime (cloud or edge).
Instead of transferring data into Ataccama ONE for processing, ONE generates SQL queries that run directly within the source, leveraging its computational resources. Only results and invalid samples (if configured) are returned to ONE.
Pushdown is supported for profiling and DQ evaluation jobs.
- Pushdown profiling
-
Profiling runs on pushdown for all supported source connections by default. There are no limitations when running profiling in pushdown compared to profiling in ONE.
- Pushdown for DQ evaluation
-
Pushdown processing for DQ evaluation can be enabled or disabled as required, as some limitations exist. See When to Use Pushdown for DQ Evaluation.
Why use pushdown processing
Using pushdown processing can improve performance, security, and scalability.
-
Performance: Evaluate millions of records in seconds.
-
Security: Data stays in your source and only results (and invalid samples, if configured) are returned. There is no data transfer to external services; all processing runs on your secured and governed source.
-
Infrastructure: Use your existing infrastructure to evaluate data—no need to set up and manage new processing servers.
Next steps
-
To decide whether pushdown is the right choice for DQ evaluation, see When to Use Pushdown for DQ Evaluation.
-
Ready to enable pushdown for DQ evaluation? See the connection page for your source:
Was this page useful?