mirror of
https://github.com/ClickHouse/ClickHouse.git
synced 2024-12-11 08:52:06 +00:00
df17cd467b
* Release more num_streams if data is small Besides the sum_marks and min_marks_for_concurrent_read, we could also involve the system cores to get the num_streams if the data is small. Increasing the num_streams and decreasing the min_marks_for_concurrent_read would improve the parallel performance if the system has plentiful cores. Test the patch on 2x80 vCPUs system. Q39 of clickbench has got 3.3x performance improvement. Q36 has got 2.6x performance improvement. The overall geomean has got 9% gain. Signed-off-by: Jiebin Sun <jiebin.sun@intel.com> * Release more num_streams if data is small Change the min marks from 4 to 8 as the profit is small and 8 granules is the default block size. Signed-off-by: Jiebin Sun <jiebin.sun@intel.com> --------- Signed-off-by: Jiebin Sun <jiebin.sun@intel.com> |
||
|---|---|---|
| .. | ||
| examples | ||
| Executors | ||
| Formats | ||
| Merges | ||
| QueryPlan | ||
| Sinks | ||
| Sources | ||
| tests | ||
| Transforms | ||
| TTL | ||
| Chunk.cpp | ||
| Chunk.h | ||
| CMakeLists.txt | ||
| ConcatProcessor.cpp | ||
| ConcatProcessor.h | ||
| DelayedPortsProcessor.cpp | ||
| DelayedPortsProcessor.h | ||
| ForkProcessor.cpp | ||
| ForkProcessor.h | ||
| IAccumulatingTransform.cpp | ||
| IAccumulatingTransform.h | ||
| IInflatingTransform.cpp | ||
| IInflatingTransform.h | ||
| IProcessor.cpp | ||
| IProcessor.h | ||
| ISimpleTransform.cpp | ||
| ISimpleTransform.h | ||
| ISink.cpp | ||
| ISink.h | ||
| ISource.cpp | ||
| ISource.h | ||
| LimitTransform.cpp | ||
| LimitTransform.h | ||
| OffsetTransform.cpp | ||
| OffsetTransform.h | ||
| PingPongProcessor.cpp | ||
| PingPongProcessor.h | ||
| Port.cpp | ||
| Port.h | ||
| QueueBuffer.h | ||
| ResizeProcessor.cpp | ||
| ResizeProcessor.h | ||
| RowsBeforeLimitCounter.h | ||