add_hash_on_top

datafusion_physical_optimizer::enforce_distribution

Function add_hash_on_top 

Source 
fn add_hash_on_top(
    input: DistributionContext,
    hash_exprs: Vec<Arc<dyn PhysicalExpr>>,
    n_target: usize,
) -> Result<DistributionContext>
Expand description

Adds a hash repartition operator:

  • to increase parallelism, and/or
  • to satisfy requirements of the subsequent operators.

Repartition(Hash) is added on top of operator input.

§Arguments

  • input: Current node.
  • hash_exprs: Stores Physical Exprs that are used during hashing.
  • n_target: desired target partition number, if partition number of the current executor is less than this value. Partition number will be increased.

§Returns

A Result object that contains new execution plan where the desired distribution is satisfied by adding a Hash repartition.