UBarney opened a new issue, #16328:
URL: https://github.com/apache/datafusion/issues/16328
### Is your feature request related to a problem or challenge?
The `NestedLoopJoinExec` operator can produce output batches with fewer rows
than the configured batch_size. To improve performance, we should add a
`CoalesceBatchesExec` operator after it.
When running the following SQL query, the metrics show that the NLJ operator
is producing very small batches: (NLJ "output_rows=6" "output_batches=3")
```
explain analyze select count(*) from range(2,8194) as t1 join range(2,24576)
as t2 on t1.value * t2.value < 10;
+-------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| plan_type | plan
|
+-------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| Plan with Metrics | ProjectionExec: expr=[count(Int64(1))@0 as count(*)],
metrics=[output_rows=1, elapsed_compute=525ns]
|
| | AggregateExec: mode=Final, gby=[],
aggr=[count(Int64(1))], metrics=[output_rows=1, elapsed_compute=17.663µs]
|
| | CoalescePartitionsExec, metrics=[output_rows=24,
elapsed_compute=17.979µs]
|
| | AggregateExec: mode=Partial, gby=[],
aggr=[count(Int64(1))], metrics=[output_rows=24, elapsed_compute=2.412497ms]
|
| | ProjectionExec: expr=[],
metrics=[output_rows=6, elapsed_compute=1.17µs]
|
| | NestedLoopJoinExec: join_type=Inner,
filter=value@0 * value@1 < 10, metrics=[output_rows=6, build_input_batches=1,
build_input_rows=8192, input_batches=3, input_rows=24574, output_batches=3,
build_mem_used=65632, build_time=32.303µs, join_time=2.490000275s] |
| | LazyMemoryExec: partitions=1,
batch_generators=[range: start=2, end=8194, batch_size=8192], metrics=[]
|
| | RepartitionExec:
partitioning=RoundRobinBatch(24), input_partitions=1,
metrics=[fetch_time=42.542µs, repartition_time=1ns, send_time=396.529µs]
|
| | LazyMemoryExec: partitions=1,
batch_generators=[range: start=2, end=24576, batch_size=8192], metrics=[]
|
| |
|
+-------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row(s) fetched.
Elapsed 0.834 seconds.
```
### Describe the solution you'd like
_No response_
### Describe alternatives you've considered
_No response_
### Additional context
_No response_
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org
For additional commands, e-mail: github-h...@datafusion.apache.org
