You are viewing a plain text version of this content. The canonical link for it is here.

Posted to notifications@asterixdb.apache.org by "Xikui Wang (Code Review)" <do...@asterixdb.incubator.apache.org> on 2017/09/07 01:03:59 UTC

Change in asterixdb[master]: [HYR] Remove table size from InMemoryHashJoin interface

Xikui Wang has uploaded a new change for review.

  https://asterix-gerrit.ics.uci.edu/1994

Change subject: [HYR] Remove table size from InMemoryHashJoin interface
......................................................................

[HYR] Remove table size from InMemoryHashJoin interface

- user model changes: no
- storage format changes: no
- interface changes: yes

Details:
As this info is available in HashTable already, table size should not
be part of the interface.

Change-Id: I02a677ecfef80ccd4332447a6dcb5d480be6fe80
---
M hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/HybridHashJoinOperatorDescriptor.java
M hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoin.java
M hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoinOperatorDescriptor.java
M hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/ISerializableTable.java
M hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/SerializableHashTable.java
5 files changed, 27 insertions(+), 36 deletions(-)


  git pull ssh://asterix-gerrit.ics.uci.edu:29418/asterixdb refs/changes/94/1994/1

diff --git a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/HybridHashJoinOperatorDescriptor.java b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/HybridHashJoinOperatorDescriptor.java
index d9e6180..665bb2b 100644
--- a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/HybridHashJoinOperatorDescriptor.java
+++ b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/HybridHashJoinOperatorDescriptor.java
@@ -309,10 +309,10 @@
                             .createPartitioner();
                     int tableSize = (int) (state.memoryForHashtable * recordsPerFrame * factor);
                     ISerializableTable table = new SimpleSerializableHashTable(tableSize, ctx);
-                    state.joiner = new InMemoryHashJoin(ctx, tableSize, new FrameTupleAccessor(rd0), hpc0,
-                            new FrameTupleAccessor(rd1), rd1, hpc1,
-                            new FrameTuplePairComparator(keys0, keys1, comparators), isLeftOuter, nullWriters1, table,
-                            predEvaluator, null);
+                    state.joiner =
+                            new InMemoryHashJoin(ctx, new FrameTupleAccessor(rd0), hpc0, new FrameTupleAccessor(rd1),
+                                    rd1, hpc1, new FrameTuplePairComparator(keys0, keys1, comparators), isLeftOuter,
+                                    nullWriters1, table, predEvaluator, null);
                     bufferForPartitions = new IFrame[state.nPartitions];
                     state.fWriters = new RunFileWriter[state.nPartitions];
                     for (int i = 0; i < state.nPartitions; i++) {
@@ -505,10 +505,11 @@
                                     continue;
                                 }
                                 table.reset();
-                                InMemoryHashJoin joiner = new InMemoryHashJoin(ctx, tableSize,
-                                        new FrameTupleAccessor(rd0), hpcRep0, new FrameTupleAccessor(rd1), rd1, hpcRep1,
-                                        new FrameTuplePairComparator(keys0, keys1, comparators), isLeftOuter,
-                                        nullWriters1, table, predEvaluator, null);
+                                InMemoryHashJoin joiner =
+                                        new InMemoryHashJoin(ctx, new FrameTupleAccessor(rd0), hpcRep0,
+                                                new FrameTupleAccessor(rd1), rd1, hpcRep1,
+                                                new FrameTuplePairComparator(keys0, keys1, comparators), isLeftOuter,
+                                                nullWriters1, table, predEvaluator, null);
 
                                 if (buildWriter != null) {
                                     RunFileReader buildReader = buildWriter.createDeleteOnCloseReader();
diff --git a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoin.java b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoin.java
index ec1c3a9..7f34d13 100644
--- a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoin.java
+++ b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoin.java
@@ -56,7 +56,6 @@
     private final boolean isLeftOuter;
     private final ArrayTupleBuilder missingTupleBuild;
     private final ISerializableTable table;
-    private final int tableSize;
     private final TuplePointer storedTuplePointer;
     private final boolean reverseOutputOrder; //Should we reverse the order of tuples, we are writing in output
     private final IPredicateEvaluator predEvaluator;
@@ -67,23 +66,22 @@
 
     private static final Logger LOGGER = Logger.getLogger(InMemoryHashJoin.class.getName());
 
-    public InMemoryHashJoin(IHyracksTaskContext ctx, int tableSize, FrameTupleAccessor accessorProbe,
+    public InMemoryHashJoin(IHyracksTaskContext ctx, FrameTupleAccessor accessorProbe,
             ITuplePartitionComputer tpcProbe, FrameTupleAccessor accessorBuild, RecordDescriptor rDBuild,
             ITuplePartitionComputer tpcBuild, FrameTuplePairComparator comparator, boolean isLeftOuter,
             IMissingWriter[] missingWritersBuild, ISerializableTable table, IPredicateEvaluator predEval,
             ISimpleFrameBufferManager bufferManager)
             throws HyracksDataException {
-        this(ctx, tableSize, accessorProbe, tpcProbe, accessorBuild, rDBuild, tpcBuild, comparator, isLeftOuter,
+        this(ctx, accessorProbe, tpcProbe, accessorBuild, rDBuild, tpcBuild, comparator, isLeftOuter,
                 missingWritersBuild, table, predEval, false, bufferManager);
     }
 
-    public InMemoryHashJoin(IHyracksTaskContext ctx, int tableSize, FrameTupleAccessor accessorProbe,
+    public InMemoryHashJoin(IHyracksTaskContext ctx, FrameTupleAccessor accessorProbe,
             ITuplePartitionComputer tpcProbe, FrameTupleAccessor accessorBuild,
             RecordDescriptor rDBuild, ITuplePartitionComputer tpcBuild, FrameTuplePairComparator comparator,
             boolean isLeftOuter, IMissingWriter[] missingWritersBuild, ISerializableTable table,
             IPredicateEvaluator predEval, boolean reverse, ISimpleFrameBufferManager bufferManager)
             throws HyracksDataException {
-        this.tableSize = tableSize;
         this.table = table;
         storedTuplePointer = new TuplePointer();
         buffers = new ArrayList<>();
@@ -109,12 +107,12 @@
         reverseOutputOrder = reverse;
         this.tupleAccessor = new TupleInFrameListAccessor(rDBuild, buffers);
         this.bufferManager = bufferManager;
-        if (tableSize != 0) {
+        if (table.getTableSize() != 0) {
             isTableCapacityNotZero = true;
         } else {
             isTableCapacityNotZero = false;
         }
-        LOGGER.fine("InMemoryHashJoin has been created for a table size of " + tableSize + " for Thread ID "
+        LOGGER.fine("InMemoryHashJoin has been created for a table size of " + table.getTableSize() + " for Thread ID "
                 + Thread.currentThread().getId() + ".");
     }
 
@@ -124,7 +122,7 @@
         accessorBuild.reset(buffer);
         int tCount = accessorBuild.getTupleCount();
         for (int i = 0; i < tCount; ++i) {
-            int entry = tpcBuild.partition(accessorBuild, i, tableSize);
+            int entry = tpcBuild.partition(accessorBuild, i, table.getTableSize());
             storedTuplePointer.reset(bIndex, i);
             // If an insertion fails, then tries to insert the same tuple pointer again after compacting the table.
             if (!table.insert(entry, storedTuplePointer)) {
@@ -160,7 +158,7 @@
     void join(int tid, IFrameWriter writer) throws HyracksDataException {
         boolean matchFound = false;
         if (isTableCapacityNotZero) {
-            int entry = tpcProbe.partition(accessorProbe, tid, tableSize);
+            int entry = tpcProbe.partition(accessorProbe, tid, table.getTableSize());
             int tupleCount = table.getTupleCount(entry);
             for (int i = 0; i < tupleCount; i++) {
                 table.getTuplePointer(entry, i, storedTuplePointer);
diff --git a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoinOperatorDescriptor.java b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoinOperatorDescriptor.java
index 702dae6..d81d955 100644
--- a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoinOperatorDescriptor.java
+++ b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/join/InMemoryHashJoinOperatorDescriptor.java
@@ -104,21 +104,6 @@
         this.memSizeInFrames = memSizeInFrames;
     }
 
-    public InMemoryHashJoinOperatorDescriptor(IOperatorDescriptorRegistry spec, int[] keys0, int[] keys1,
-            IBinaryHashFunctionFactory[] hashFunctionFactories, IBinaryComparatorFactory[] comparatorFactories,
-            RecordDescriptor recordDescriptor, int tableSize, int memSizeInFrames) {
-        this(spec, keys0, keys1, hashFunctionFactories, comparatorFactories, recordDescriptor, tableSize, null,
-                memSizeInFrames);
-    }
-
-    public InMemoryHashJoinOperatorDescriptor(IOperatorDescriptorRegistry spec, int[] keys0, int[] keys1,
-            IBinaryHashFunctionFactory[] hashFunctionFactories, IBinaryComparatorFactory[] comparatorFactories,
-            RecordDescriptor recordDescriptor, boolean isLeftOuter, IMissingWriterFactory[] nullWriterFactories1,
-            int tableSize, int memSizeInFrames) {
-        this(spec, keys0, keys1, hashFunctionFactories, comparatorFactories, null, recordDescriptor, isLeftOuter,
-                nullWriterFactories1, tableSize, memSizeInFrames);
-    }
-
     @Override
     public void contributeActivities(IActivityGraphBuilder builder) {
         ActivityId hbaId = new ActivityId(odId, 0);
@@ -203,10 +188,10 @@
                     state = new HashBuildTaskState(ctx.getJobletContext().getJobId(),
                             new TaskId(getActivityId(), partition));
                     ISerializableTable table = new SerializableHashTable(tableSize, ctx, bufferManager);
-                    state.joiner = new InMemoryHashJoin(ctx, tableSize, new FrameTupleAccessor(rd0), hpc0,
-                            new FrameTupleAccessor(rd1), rd1, hpc1,
-                            new FrameTuplePairComparator(keys0, keys1, comparators), isLeftOuter, nullWriters1, table,
-                            predEvaluator, bufferManager);
+                    state.joiner =
+                            new InMemoryHashJoin(ctx, new FrameTupleAccessor(rd0), hpc0, new FrameTupleAccessor(rd1),
+                                    rd1, hpc1, new FrameTuplePairComparator(keys0, keys1, comparators), isLeftOuter,
+                                    nullWriters1, table, predEvaluator, bufferManager);
                 }
 
                 @Override
diff --git a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/ISerializableTable.java b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/ISerializableTable.java
index d0e0616..dea9c98 100644
--- a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/ISerializableTable.java
+++ b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/ISerializableTable.java
@@ -64,4 +64,6 @@
      * Prints out the internal information of this table.
      */
     String printInfo();
+
+    int getTableSize();
 }
diff --git a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/SerializableHashTable.java b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/SerializableHashTable.java
index ca97be3..3910e41 100644
--- a/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/SerializableHashTable.java
+++ b/hyracks-fullstack/hyracks/hyracks-dataflow-std/src/main/java/org/apache/hyracks/dataflow/std/structures/SerializableHashTable.java
@@ -566,4 +566,9 @@
             this.second = second;
         }
     }
+
+    @Override
+    public int getTableSize() {
+        return tableSize;
+    }
 }

-- 
To view, visit https://asterix-gerrit.ics.uci.edu/1994
To unsubscribe, visit https://asterix-gerrit.ics.uci.edu/settings

Gerrit-MessageType: newchange
Gerrit-Change-Id: I02a677ecfef80ccd4332447a6dcb5d480be6fe80
Gerrit-PatchSet: 1
Gerrit-Project: asterixdb
Gerrit-Branch: master
Gerrit-Owner: Xikui Wang <xk...@gmail.com>