Use node-level local limit (#20)

* Use node-level local limit

* serialize limit in shuffle writer

* Revert "Merge pull request #19 from coralogix/sc-5792"

This reverts commit 08140ef, reversing
changes made to a7f1384.

* add log

* make sure we don't forget limit for shuffle writer

* update accum correctly and try to break early

* Check local limit accumulator before polling for more data

* fix build

Co-authored-by: Martins Purins <[email protected]>

Author: 2 people

ballista/core/Cargo.toml

@@ -56,9 +56,11 @@ datafusion-proto = "17.0.0"
 futures = "0.3"
 hashbrown = "0.13"
 
+lazy_static = "1.4.0"
 itertools = "0.10"
 libloading = "0.7.3"
 log = "0.4"
+lru = "0.8.1"
 object_store = "0.5.2"
 once_cell = "1.9.0"
 

ballista/core/proto/ballista.proto

@@ -53,6 +53,134 @@ message UnresolvedShuffleExecNode {
   uint32 output_partition_count = 4;
 }
 
+message FilterExecNode {
+  PhysicalPlanNode input = 1;
+  PhysicalExprNode expr = 2;
+}
+
+message FileGroup {
+  repeated PartitionedFile files = 1;
+}
+
+message ScanLimit {
+  // wrap into a message to make it optional
+  uint32 limit = 1;
+}
+
+message FileScanExecConf {
+  repeated FileGroup file_groups = 1;
+  datafusion.Schema schema = 2;
+  repeated uint32 projection = 4;
+  ScanLimit limit = 5;
+  Statistics statistics = 6;
+  repeated string table_partition_cols = 7;
+  string object_store_url = 8;
+}
+
+message ParquetScanExecNode {
+  FileScanExecConf base_conf = 1;
+  datafusion.LogicalExprNode pruning_predicate = 2;
+}
+
+message CsvScanExecNode {
+  FileScanExecConf base_conf = 1;
+  bool has_header = 2;
+  string delimiter = 3;
+}
+
+message AvroScanExecNode {
+  FileScanExecConf base_conf = 1;
+}
+
+enum PartitionMode {
+  COLLECT_LEFT = 0;
+  PARTITIONED = 1;
+}
+
+message HashJoinExecNode {
+  PhysicalPlanNode left = 1;
+  PhysicalPlanNode right = 2;
+  repeated JoinOn on = 3;
+  datafusion.JoinType join_type = 4;
+  PartitionMode partition_mode = 6;
+  bool null_equals_null = 7;
+  JoinFilter filter = 8;
+}
+
+message UnionExecNode {
+  repeated PhysicalPlanNode inputs = 1;
+}
+
+message ExplainExecNode {
+  datafusion.Schema schema = 1;
+  repeated datafusion.StringifiedPlan stringified_plans = 2;
+  bool verbose = 3;
+}
+
+message CrossJoinExecNode {
+  PhysicalPlanNode left = 1;
+  PhysicalPlanNode right = 2;
+}
+
+message PhysicalColumn {
+  string name = 1;
+  uint32 index = 2;
+}
+
+message JoinOn {
+  PhysicalColumn left = 1;
+  PhysicalColumn right = 2;
+}
+
+message EmptyExecNode {
+  bool produce_one_row = 1;
+  datafusion.Schema schema = 2;
+}
+
+message ProjectionExecNode {
+  PhysicalPlanNode input = 1;
+  repeated PhysicalExprNode expr = 2;
+  repeated string expr_name = 3;
+}
+
+enum AggregateMode {
+  PARTIAL = 0;
+  FINAL = 1;
+  FINAL_PARTITIONED = 2;
+}
+
+message WindowAggExecNode {
+  PhysicalPlanNode input = 1;
+  repeated PhysicalExprNode window_expr = 2;
+  repeated string window_expr_name = 3;
+  datafusion.Schema input_schema = 4;
+}
+
+message AggregateExecNode {
+  repeated PhysicalExprNode group_expr = 1;
+  repeated PhysicalExprNode aggr_expr = 2;
+  AggregateMode mode = 3;
+  PhysicalPlanNode input = 4;
+  repeated string group_expr_name = 5;
+  repeated string aggr_expr_name = 6;
+  // we need the input schema to the partial aggregate to pass to the final aggregate
+  datafusion.Schema input_schema = 7;
+  repeated PhysicalExprNode null_expr = 8;
+  repeated bool groups = 9;
+}
+
+message ShuffleWriterExecNode {
+  //TODO it seems redundant to provide job and stage id here since we also have them
+  // in the TaskDefinition that wraps this plan
+  string job_id = 1;
+  uint32 stage_id = 2;
+  PhysicalPlanNode input = 3;
+  PhysicalHashRepartition output_partitioning = 4;
+  oneof optional_limit {
+    uint64 limit = 6;
+  }
+}
+
 message ShuffleReaderExecNode {
   repeated ShuffleReaderPartition partition = 1;
   datafusion.Schema schema = 2;

ballista/core/src/execution_plans/shuffle_writer.rs

@@ -55,7 +55,30 @@ use datafusion::arrow::error::ArrowError;
 use datafusion::execution::context::TaskContext;
 use datafusion::physical_plan::repartition::BatchPartitioner;
 use datafusion::physical_plan::stream::RecordBatchStreamAdapter;
+use lazy_static::lazy_static;
 use log::{debug, info};
+use lru::LruCache;
+use parking_lot::Mutex;
+use std::num::NonZeroUsize;
+use std::sync::atomic::{AtomicUsize, Ordering};
+
+lazy_static! {
+    static ref LIMIT_ACCUMULATORS: Mutex<LruCache<(String, usize), Arc<AtomicUsize>>> =
+        Mutex::new(LruCache::new(NonZeroUsize::new(40).unwrap()));
+}
+
+fn get_limit_accumulator(job_id: &str, stage: usize) -> Arc<AtomicUsize> {
+    let mut guard = LIMIT_ACCUMULATORS.lock();
+
+    if let Some(accumulator) = guard.get(&(job_id.to_owned(), stage)) {
+        accumulator.clone()
+    } else {
+        let accumulator = Arc::new(AtomicUsize::new(0));
+        guard.push((job_id.to_owned(), stage), accumulator.clone());
+
+        accumulator
+    }
+}
 
 /// ShuffleWriterExec represents a section of a query plan that has consistent partitioning and
 /// can be executed as one unit with each partition being executed in parallel. The output of each
@@ -75,6 +98,8 @@ pub struct ShuffleWriterExec {
     shuffle_output_partitioning: Option<Partitioning>,
     /// Execution metrics
     metrics: ExecutionPlanMetricsSet,
+    /// Maximum number of rows to return
+    limit: Option<usize>,
 }
 
 #[derive(Debug, Clone)]
@@ -121,6 +146,26 @@ impl ShuffleWriterExec {
             work_dir,
             shuffle_output_partitioning,
             metrics: ExecutionPlanMetricsSet::new(),
+            limit: None,
+        })
+    }
+
+    pub fn try_new_with_limit(
+        job_id: String,
+        stage_id: usize,
+        plan: Arc<dyn ExecutionPlan>,
+        work_dir: String,
+        shuffle_output_partitioning: Option<Partitioning>,
+        limit: Option<usize>,
+    ) -> Result<Self> {
+        Ok(Self {
+            job_id,
+            stage_id,
+            plan,
+            work_dir,
+            shuffle_output_partitioning,
+            metrics: ExecutionPlanMetricsSet::new(),
+            limit,
         })
     }
 
@@ -139,6 +184,10 @@ impl ShuffleWriterExec {
         self.shuffle_output_partitioning.as_ref()
     }
 
+    pub fn limit(&self) -> Option<usize> {
+        self.limit
+    }
+
     pub fn execute_shuffle_write(
         &self,
         input_partition: usize,
@@ -152,6 +201,10 @@ impl ShuffleWriterExec {
         let output_partitioning = self.shuffle_output_partitioning.clone();
         let plan = self.plan.clone();
 
+        let limit_and_accumulator = self
+            .limit
+            .map(|l| (l, get_limit_accumulator(&self.job_id, self.stage_id)));
+
         async move {
             let now = Instant::now();
             let mut stream = plan.execute(input_partition, context)?;
@@ -170,6 +223,7 @@ impl ShuffleWriterExec {
                         &mut stream,
                         path,
                         &write_metrics.write_time,
+                        limit_and_accumulator,
                     )
                     .await
                     .map_err(|e| DataFusionError::Execution(format!("{e:?}")))?;
@@ -211,10 +265,26 @@ impl ShuffleWriterExec {
                         write_metrics.repart_time.clone(),
                     )?;
 
-                    while let Some(result) = stream.next().await {
+                    while let Some(result) = {
+                        let poll_more = limit_and_accumulator.as_ref().map_or(
+                            true,
+                            |(limit, accum)| {
+                                let total_rows = accum.load(Ordering::SeqCst);
+                                total_rows < *limit
+                            },
+                        );
+
+                        if poll_more {
+                            stream.next().await
+                        } else {
+                            None
+                        }
+                    } {
                         let input_batch = result?;
 
-                        write_metrics.input_rows.add(input_batch.num_rows());
+                        let num_rows = input_batch.num_rows();
+
+                        write_metrics.input_rows.add(num_rows);
 
                         partitioner.partition(
                             input_batch,
@@ -252,6 +322,13 @@ impl ShuffleWriterExec {
                                 Ok(())
                             },
                         )?;
+
+                        if let Some((limit, accum)) = limit_and_accumulator.as_ref() {
+                            let total_rows = accum.fetch_add(num_rows, Ordering::SeqCst);
+                            if total_rows > *limit {
+                                break;
+                            }
+                        }
                     }
 
                     let mut part_locs = vec![];
@@ -320,12 +397,13 @@ impl ExecutionPlan for ShuffleWriterExec {
         self: Arc<Self>,
         children: Vec<Arc<dyn ExecutionPlan>>,
     ) -> Result<Arc<dyn ExecutionPlan>> {
-        Ok(Arc::new(ShuffleWriterExec::try_new(
+        Ok(Arc::new(ShuffleWriterExec::try_new_with_limit(
             self.job_id.clone(),
             self.stage_id,
             children[0].clone(),
             self.work_dir.clone(),
             self.shuffle_output_partitioning.clone(),
+            self.limit,
         )?))
     }