* Monomorphic processing of topN queries with simple double aggregators and historical segments
* Add CalledFromHotLoop annocations to specialized methods in SimpleDoubleBufferAggregator
* Fix a bug in Historical1SimpleDoubleAggPooledTopNScannerPrototype
* Fix a bug in SpecializationService
* In SpecializationService, emit maxSpecializations warning only once
* Make GenericIndexed.theBuffer final
* Address comments
* Newline
* Reapply 439c906 (Make GenericIndexed.theBuffer final)
* Remove extra PooledTopNAlgorithm.capabilities field
* Improve CachingIndexed.inspectRuntimeShape()
* Fix CompressedVSizeIntsIndexedSupplier.inspectRuntimeShape()
* Don't override inspectRuntimeShape() in subclasses of CompressedVSizeIndexedInts
* Annotate methods in specializations of DimensionSelector and FloatColumnSelector with @CalledFromHotLoop
* Make ValueMatcher to implement HotLoopCallee
* Doc fix
* Fix inspectRuntimeShape() impl in ExpressionSelectors
* INFO logging of specialization events
* Remove modificator
* Fix OrFilter
* Fix AndFilter
* Refactor PooledTopNAlgorithm.scanAndAggregate()
* Small refactoring
* Add 'nothing to inspect' messages in empty HotLoopCallee.inspectRuntimeShape() implementations
* Don't care about runtime shape in tests
* Fix accessor bugs in Historical1SimpleDoubleAggPooledTopNScannerPrototype and HistoricalSingleValueDimSelector1SimpleDoubleAggPooledTopNScannerPrototype, cover them with tests
* Doc wording
* Address comments
* Remove MagicAccessorBridge and ensure Offset subclasses are public
* Attach error message to element
* Monomorphic processing: add HotLoopCallee, CalledFromHotLoop, RuntimeShapeInspector, SpecializationService. Specialize topN queries with 1 or 2 aggregators. Add Cursor.advanceUninterruptibly() and isDoneOrInterrupted() for exception-free query processing.
* Use Execs.singleThreaded()
* RuntimeShapeInspector to support nullable fields
* Make CalledFromHotLoop annotation Inherited
* Remove unnecessary conversion of array of ColumnSelectorPluses to list and back to array in CardinalityAggregatorFactory
* Close InputStream in SpecializationService
* Formatting
* Test specialized PooledTopNScanners
* Set flags in PooledTopNAlgorithm directly
* Fix tests, dependent on CountAggragatorFactory toString() form
* Fix
* Revert CountAggregatorFactory changes
* Implement inspectRuntimeShape() for LongWrappingDimensionSelector and FloatWrappingDimensionSelector
* Remove duplicate RoaringBitmap dependency in the extendedset pom.xml
* Fix
* Treat ByteBuffers specially in StringRuntimeShape
* Doc fix
* Annotate BufferAggregator.init() with CalledFromHotLoop
* Make triggerSpecializationIterationsThreshold an int
* Remove SpecializationService.PerPrototypeClassState.of()
* Add comments
* Limit the amount of specializations that SpecializationService could make
* Add default implementation for BufferAggregator.inspectRuntimeShape(), for compatibility with extensions
* Use more efficient ConcurrentMap's idioms in SpecializationService
* No more singleton. Reduce iterations
* Granularities
* Fix the delay in the test
* Add license header
* Remove unused imports
* Lot more unused imports from all the rearranging
* CR feedback
* Move javadoc to constructor
* Refactor Segment Granularity
* Beginning of one granularity
* Copy the fix for custom periods in segment-grunalrity over here.
* Remove the custom serialization for now.
* Compilation cleanup
* Reformat code
* Fixing unit tests
* Unify to use a single iterable
* Backward compatibility for rolling upgrade
* Minor check style. Cosmetic changes.
* Rename length and millis to duration
* CR feedback
* Minor changes.
This patch introduces a GroupByStrategy concept and two strategies: "v1"
is the current groupBy strategy and "v2" is a new one. It also introduces
a merge buffers concept in DruidProcessingModule, to try to better
manage memory used for merging.
Both of these are described in more detail in #2987.
There are two goals of this patch:
1. Make it possible for historical/realtime nodes to return larger groupBy
result sets, faster, with better memory management.
2. Make it possible for brokers to merge streams when there are no order-by
columns, avoiding materialization.
This patch does not do anything to help with memory management on the broker
when there are order-by columns or when there are nested queries. That could
potentially be done in a future patch.
Roman Leventov
Benedict Jin
Gian Merlino
Roman Leventov
praveev
Jonathan Wei
Akash Dwivedi
Charles Allen
fjy
binlijin