Skip to content

Consolidate sort and external_sort #1596

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 9 commits into from
Jan 21, 2022
Merged

Consolidate sort and external_sort #1596

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
08d0dc3
Change SPMS to use heap sort, use SPMS instead of in-mem-sort as well
yjshen Jan 17, 2022
4c099f7
Incorporate metrics, external_sort pass all sort tests
yjshen Jan 17, 2022
419a362
Remove the original sort, substitute with external sort
yjshen Jan 18, 2022
ae3030c
Merge remote-tracking branch 'apache/master' into consol_ims_spms
yjshen Jan 18, 2022
58817bb
Fix different batch_size setting in SPMS test
yjshen Jan 18, 2022
dd9886d
Change to use combine and sort for in memory N-way merge
yjshen Jan 19, 2022
d28b96c
Resolve comments on async and doc
yjshen Jan 19, 2022
5b35c2a
Update sort to avoid deadlock during spilling
yjshen Jan 19, 2022
6a458c6
Fix spill hanging
yjshen Jan 20, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions datafusion/src/execution/context.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1210,6 +1210,7 @@ impl FunctionRegistry for ExecutionContextState {
#[cfg(test)]
mod tests {
use super::*;
use crate::execution::context::QueryPlanner;
use crate::from_slice::FromSlice;
use crate::logical_plan::plan::Projection;
use crate::logical_plan::TableScan;
Expand Down
14 changes: 11 additions & 3 deletions datafusion/src/physical_plan/common.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,7 @@
use super::{RecordBatchStream, SendableRecordBatchStream};
use crate::error::{DataFusionError, Result};
use crate::execution::runtime_env::RuntimeEnv;
use crate::physical_plan::metrics::BaselineMetrics;
use crate::physical_plan::{ColumnStatistics, ExecutionPlan, Statistics};
use arrow::compute::concat;
use arrow::datatypes::{Schema, SchemaRef};
Expand All @@ -41,15 +42,21 @@ pub struct SizedRecordBatchStream {
schema: SchemaRef,
batches: Vec<Arc<RecordBatch>>,
index: usize,
baseline_metrics: BaselineMetrics,
}

impl SizedRecordBatchStream {
/// Create a new RecordBatchIterator
pub fn new(schema: SchemaRef, batches: Vec<Arc<RecordBatch>>) -> Self {
pub fn new(
schema: SchemaRef,
batches: Vec<Arc<RecordBatch>>,
baseline_metrics: BaselineMetrics,
) -> Self {
SizedRecordBatchStream {
schema,
index: 0,
batches,
baseline_metrics,
}
}
}
Expand All @@ -61,12 +68,13 @@ impl Stream for SizedRecordBatchStream {
mut self: std::pin::Pin<&mut Self>,
_: &mut Context<'_>,
) -> Poll<Option<Self::Item>> {
Poll::Ready(if self.index < self.batches.len() {
let poll = Poll::Ready(if self.index < self.batches.len() {
self.index += 1;
Some(Ok(self.batches[self.index - 1].as_ref().clone()))
} else {
None
})
});
self.baseline_metrics.record_poll(poll)
Copy link
Member Author

@yjshen yjshen Jan 17, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Make metrics right when there is only one record batch.

}
}

Expand Down
5 changes: 5 additions & 0 deletions datafusion/src/physical_plan/explain.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -32,6 +32,7 @@ use arrow::{array::StringBuilder, datatypes::SchemaRef, record_batch::RecordBatc

use super::SendableRecordBatchStream;
use crate::execution::runtime_env::RuntimeEnv;
use crate::physical_plan::metrics::{BaselineMetrics, ExecutionPlanMetricsSet};
use async_trait::async_trait;

/// Explain execution plan operator. This operator contains the string
Expand Down Expand Up @@ -146,9 +147,13 @@ impl ExecutionPlan for ExplainExec {
],
)?;

let metrics = ExecutionPlanMetricsSet::new();
let baseline_metrics = BaselineMetrics::new(&metrics, partition);

Ok(Box::pin(SizedRecordBatchStream::new(
self.schema.clone(),
vec![Arc::new(record_batch)],
baseline_metrics,
)))
}

Expand Down
Loading