V1.1.0
Change Description
The OmniData operator pushdown feature of Kunpeng BoostKit for Big Data pushes operators on compute nodes to storage nodes and transmits the computing results to compute nodes, reducing the amount of data transmitted over the network.
Version 1.1.0 has the following changes:
- Supports connection to the homogeneous acceleration framework (HAF) to replace the server/client interfaces of the original gRPC communication pushdown framework and implements pushdown using annotations.
- Supports operator pushdown to Ceph storage nodes for processing.
Resolved Issues
- Added logs to facilitate fault locating.
- Fixed data type inconsistency issues due to Cast conversion in Spark statements.
Known Issues
None
Parent topic: OmniData