Merge a77a362 into a7a0baa

apache · Mar 26, 2018 · 8046552 · 8046552
2 parents a7a0baa + a77a362
commit 8046552
Show file tree

Hide file tree

Showing 31 changed files with 447 additions and 1,609 deletions.
diff --git a/...cube/src/main/java/org/apache/kylin/cube/cuboid/algorithm/AbstractRecommendAlgorithm.java b/...cube/src/main/java/org/apache/kylin/cube/cuboid/algorithm/AbstractRecommendAlgorithm.java
@@ -18,16 +18,17 @@
 
 package org.apache.kylin.cube.cuboid.algorithm;
 
-import java.util.concurrent.atomic.AtomicBoolean;
-
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
+import java.util.List;
+import java.util.concurrent.atomic.AtomicBoolean;
+
 public abstract class AbstractRecommendAlgorithm implements CuboidRecommendAlgorithm {
     private static final Logger logger = LoggerFactory.getLogger(AbstractRecommendAlgorithm.class);
 
-    private CuboidStats cuboidStats;
-    private BenefitPolicy benefitPolicy;
+    protected final CuboidStats cuboidStats;
+    protected final BenefitPolicy benefitPolicy;
 
     private AtomicBoolean cancelRequested = new AtomicBoolean(false);
     private AtomicBoolean canceled = new AtomicBoolean(false);
@@ -44,14 +45,19 @@ public AbstractRecommendAlgorithm(final long timeout, BenefitPolicy benefitPolic
         this.benefitPolicy = benefitPolicy;
     }
 
+    @Override
+    public List<Long> recommend(double expansionRate) {
+        double spaceLimit = cuboidStats.getBaseCuboidSize() * expansionRate;
+        return start(spaceLimit);
+    }
+
     @Override
     public void cancel() {
         cancelRequested.set(true);
     }
 
     /**
      * Checks whether the algorithm has been canceled or timed out.
-     * 
      */
     protected boolean shouldCancel() {
         if (canceled.get()) {
@@ -71,12 +77,4 @@ protected boolean shouldCancel() {
         }
         return false;
     }
-
-    public CuboidStats getCuboidStats() {
-        return cuboidStats;
-    }
-
-    public BenefitPolicy getBenefitPolicy() {
-        return benefitPolicy;
-    }
 }
diff --git a/core-cube/src/main/java/org/apache/kylin/cube/cuboid/algorithm/BPUSCalculator.java b/core-cube/src/main/java/org/apache/kylin/cube/cuboid/algorithm/BPUSCalculator.java
@@ -18,15 +18,14 @@
 
 package org.apache.kylin.cube.cuboid.algorithm;
 
-import java.util.List;
-import java.util.Map;
-import java.util.Set;
-
+import com.google.common.collect.ImmutableMap;
+import com.google.common.collect.Maps;
+import com.google.common.collect.Sets;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
-import com.google.common.collect.Maps;
-import com.google.common.collect.Sets;
+import java.util.Map;
+import java.util.Set;
 
 /**
  * Calculate the benefit based on Benefit Per Unit Space.
@@ -35,17 +34,24 @@ public class BPUSCalculator implements BenefitPolicy {
 
     private static Logger logger = LoggerFactory.getLogger(BPUSCalculator.class);
 
-    protected CuboidStats cuboidStats;
-    protected Map<Long, Long> cuboidAggCostMap;
+    protected final CuboidStats cuboidStats;
+    protected final ImmutableMap<Long, Long> initCuboidAggCostMap;
+    protected final Map<Long, Long> processCuboidAggCostMap;
 
     public BPUSCalculator(CuboidStats cuboidStats) {
         this.cuboidStats = cuboidStats;
-        this.cuboidAggCostMap = Maps.newHashMap();
+        this.initCuboidAggCostMap = ImmutableMap.copyOf(initCuboidAggCostMap());
+        this.processCuboidAggCostMap = Maps.newHashMap(initCuboidAggCostMap);
     }
 
-    @Override
-    public void initBeforeStart() {
-        cuboidAggCostMap.clear();
+    protected BPUSCalculator(CuboidStats cuboidStats, ImmutableMap<Long, Long> initCuboidAggCostMap) {
+        this.cuboidStats = cuboidStats;
+        this.initCuboidAggCostMap = initCuboidAggCostMap;
+        this.processCuboidAggCostMap = Maps.newHashMap(initCuboidAggCostMap);
+    }
+
+    private Map<Long, Long> initCuboidAggCostMap() {
+        Map<Long, Long> cuboidAggCostMap = Maps.newHashMap();
         //Initialize stats for mandatory cuboids
         for (Long cuboid : cuboidStats.getAllCuboidsForMandatory()) {
             if (getCuboidCost(cuboid) != null) {
@@ -66,6 +72,7 @@ public void initBeforeStart() {
             }
             cuboidAggCostMap.put(cuboid, leastCost);
         }
+        return cuboidAggCostMap;
     }
 
     @Override
@@ -88,22 +95,21 @@ public CuboidBenefitModel.BenefitModel calculateBenefit(long cuboid, Set<Long> s
     }
 
     @Override
-    public CuboidBenefitModel.BenefitModel calculateBenefitTotal(List<Long> cuboidsToAdd, Set<Long> selected) {
+    public CuboidBenefitModel.BenefitModel calculateBenefitTotal(Set<Long> cuboidsToAdd, Set<Long> selected) {
         Set<Long> selectedInner = Sets.newHashSet(selected);
-        Map<Long, Long> cuboidAggCostMapSnapshot = Maps.newHashMap(cuboidAggCostMap);
+        Map<Long, Long> cuboidAggCostMapCopy = Maps.newHashMap(processCuboidAggCostMap);
         for (Long cuboid : cuboidsToAdd) {
             selectedInner.add(cuboid);
-            propagateAggregationCost(cuboid, selectedInner);
+            propagateAggregationCost(cuboid, selectedInner, cuboidAggCostMapCopy);
         }
         double totalCostSaving = 0;
         int benefitCount = 0;
-        for (Long cuboid : cuboidAggCostMap.keySet()) {
-            if (cuboidAggCostMap.get(cuboid) < cuboidAggCostMapSnapshot.get(cuboid)) {
-                totalCostSaving += cuboidAggCostMapSnapshot.get(cuboid) - cuboidAggCostMap.get(cuboid);
+        for (Long cuboid : cuboidAggCostMapCopy.keySet()) {
+            if (cuboidAggCostMapCopy.get(cuboid) < processCuboidAggCostMap.get(cuboid)) {
+                totalCostSaving += processCuboidAggCostMap.get(cuboid) - cuboidAggCostMapCopy.get(cuboid);
                 benefitCount++;
             }
         }
-        cuboidAggCostMap = cuboidAggCostMapSnapshot;
 
         double benefitPerUnitSpace = totalCostSaving;
         return new CuboidBenefitModel.BenefitModel(benefitPerUnitSpace, benefitCount);
@@ -120,7 +126,7 @@ protected Long getCuboidCost(long cuboid) {
     }
 
     private long getCuboidAggregationCost(long cuboid) {
-        return cuboidAggCostMap.get(cuboid);
+        return processCuboidAggCostMap.get(cuboid);
     }
 
     @Override
@@ -139,11 +145,15 @@ public double getMinBenefitRatio() {
 
     @Override
     public void propagateAggregationCost(long cuboid, Set<Long> selected) {
+        propagateAggregationCost(cuboid, selected, processCuboidAggCostMap);
+    }
+
+    public void propagateAggregationCost(long cuboid, Set<Long> selected, Map<Long, Long> processCuboidAggCostMap) {
         long aggregationCost = getCuboidCost(cuboid);
         Set<Long> childrenCuboids = cuboidStats.getAllDescendants(cuboid);
         for (Long child : childrenCuboids) {
             if (!selected.contains(child) && (aggregationCost < getCuboidAggregationCost(child))) {
-                cuboidAggCostMap.put(child, aggregationCost);
+                processCuboidAggCostMap.put(child, aggregationCost);
             }
         }
     }
@@ -158,8 +168,6 @@ public double calculateSpaceCost(long cuboid) {
 
     @Override
     public BenefitPolicy getInstance() {
-        BPUSCalculator bpusCalculator = new BPUSCalculator(this.cuboidStats);
-        bpusCalculator.cuboidAggCostMap.putAll(this.cuboidAggCostMap);
-        return bpusCalculator;
+        return new BPUSCalculator(this.cuboidStats, this.initCuboidAggCostMap);
     }
 }
diff --git a/core-cube/src/main/java/org/apache/kylin/cube/cuboid/algorithm/BenefitPolicy.java b/core-cube/src/main/java/org/apache/kylin/cube/cuboid/algorithm/BenefitPolicy.java
@@ -18,7 +18,6 @@
 
 package org.apache.kylin.cube.cuboid.algorithm;
 
-import java.util.List;
 import java.util.Set;
 
 /**
@@ -27,15 +26,29 @@
 */
 public interface BenefitPolicy {
 
+    /**
+     * @return a cloned instance with initial status
+     */
     public BenefitPolicy getInstance();
 
-    public void initBeforeStart();
-
+    /**
+     * @param selected should not be changed
+     *                 Will not change the inner instance status
+     */
     public CuboidBenefitModel.BenefitModel calculateBenefit(long cuboid, Set<Long> selected);
 
-    public CuboidBenefitModel.BenefitModel calculateBenefitTotal(List<Long> cuboidsToAdd, Set<Long> selected);
+    /**
+     * @param cuboidsToAdd should not be changed
+     * @param selected     should not be changed
+     *                     Will not change the inner instance status
+     */
+    public CuboidBenefitModel.BenefitModel calculateBenefitTotal(Set<Long> cuboidsToAdd, Set<Long> selected);
 
     public boolean ifEfficient(CuboidBenefitModel best);
 
+    /**
+     * @param selected should not be changed
+     * Will update the inner instance status
+     */
     public void propagateAggregationCost(long cuboid, Set<Long> selected);
 }