Change History

Date	Description
2026-04-01	This issue is the seventh official release. Separated the OmniData Feature Guide into an independent document from the OmniRuntime Feature Guide. Separated the OmniData Release Notes into an independent document from the OmniRuntime Release Notes.
2023-03-30	This issue is the sixth official release. The OmniData feature accelerates native computing using the Filter operator and filters Parquet type data using the column index and a bloom filter. It supports overlay of operator pushdown and operator acceleration features on the Spark engine, the OmniVec data processing format, selectivity calculation for the string type, and pushdown of some tasks.
2022-12-30	This issue is the fifth official release. Upgraded OmniData to 1.4.0 and changed the deployment mode to client/server decoupling. Optimized Hive engine performance in scenarios where the Filter operator is used. Upgraded HAF to 1.3.0 and allowed independent deployment.
2022-07-08	This issue is the fourth official release. Upgraded HAF to 1.2.0 and removed HAF user dependency by adding its support for common user deployment. Added the support for UDFs and data statistics by OmniData and upgraded the OmniData version to 1.3.0.
2022-04-08	This issue is the third official release. Upgraded the OmniData version to 1.2.0. Upgraded the HAF version to 1.0.1.
2021-12-30	This issue is the second official release. Added these capabilities: Integrates with the Homogeneous Acceleration Framework (HAF), replaces the server–client interface of the original gRPC-based pushdown framework, and enables operator pushdown to Ceph storage nodes through annotations.
2021-09-30	This issue is the first official release. The OmniData feature pushes operators on the compute side to storage nodes and transmits the procession results to compute nodes through the network, reducing the amount of data transmitted over the network.