Expand description
Structures for Morsel Driven IO.
NOTE: As of DataFusion 54.0.0, these are experimental APIs that may change substantially.
Morsel Driven IO is a technique for parallelizing the reading of large files by dividing them into smaller “morsels” that are processed independently.
It is inspired by the paper Morsel-Driven Parallelism: A NUMA-Aware Query Evaluation Framework for the Many-Core Age.
Structs§
- Morsel
Plan - Return result of
MorselPlanner::plan. - Pending
Morsel Planner - Wrapper for I/O that must complete before planning can continue.
Traits§
- Morsel
- A Morsel of work ready to resolve to a stream of
RecordBatches. - Morsel
Planner - A Morsel Planner is responsible for creating morsels for a given scan.
- Morselizer
- A Morselizer takes a single
PartitionedFileand creates the initial planner for that file.