SOLR-16989: Optimize and consolidate reuse of DocValues iterators for…

… value retrieval (#1938)

(cherry picked from commit dafbe80)

solr/CHANGES.txt

@@ -89,6 +89,7 @@ Optimizations
 
 * SOLR-16265: reduce memory usage of ContentWriter based requests in Http2SolrClient (Alex Deparvu, Kevin Risden, David Smiley)
 
+* SOLR-16989: Optimize and consolidate reuse of DocValues iterators for value retrieval (Michael Gibney)
 
 Bug Fixes
 ---------------------

solr/core/src/java/org/apache/solr/handler/component/RealTimeGetComponent.java

@@ -81,6 +81,7 @@
 import org.apache.solr.schema.IndexSchema;
 import org.apache.solr.schema.SchemaField;
 import org.apache.solr.search.DocList;
+import org.apache.solr.search.DocValuesIteratorCache;
 import org.apache.solr.search.QueryUtils;
 import org.apache.solr.search.ReturnFields;
 import org.apache.solr.search.SolrDocumentFetcher;
@@ -238,6 +239,7 @@ public void process(ResponseBuilder rb) throws IOException {
 
       boolean opennedRealtimeSearcher = false;
       BytesRefBuilder idBytes = new BytesRefBuilder();
+      DocValuesIteratorCache reuseDvIters = null;
       for (String idStr : reqIds.allIds) {
         fieldType.readableToIndexed(idStr, idBytes);
         // if _route_ is passed, id is a child doc.  TODO remove in SOLR-15064
@@ -348,7 +350,11 @@ public void process(ResponseBuilder rb) throws IOException {
             searcherInfo.getSearcher().doc(docid, rsp.getReturnFields().getLuceneFieldNames());
         SolrDocument doc = toSolrDoc(luceneDocument, core.getLatestSchema());
         SolrDocumentFetcher docFetcher = searcherInfo.getSearcher().getDocFetcher();
-        docFetcher.decorateDocValueFields(doc, docid, docFetcher.getNonStoredDVs(true));
+        if (reuseDvIters == null) {
+          reuseDvIters = new DocValuesIteratorCache(searcherInfo.getSearcher());
+        }
+        docFetcher.decorateDocValueFields(
+            doc, docid, docFetcher.getNonStoredDVs(true), reuseDvIters);
         if (null != transformer) {
           if (null == resultContext) {
             // either first pass, or we've re-opened searcher - either way now we setContext
@@ -575,7 +581,11 @@ private static SolrDocument mergePartialDocWithFullDocFromIndex(
       if (!doc.containsKey(VERSION_FIELD)) {
         searcher
             .getDocFetcher()
-            .decorateDocValueFields(doc, docid, Collections.singleton(VERSION_FIELD));
+            .decorateDocValueFields(
+                doc,
+                docid,
+                Collections.singleton(VERSION_FIELD),
+                new DocValuesIteratorCache(searcher, false));
       }
 
       long docVersion = (long) doc.getFirstValue(VERSION_FIELD);

solr/core/src/java/org/apache/solr/handler/export/BoolFieldWriter.java

@@ -21,10 +21,14 @@
 import org.apache.lucene.util.BytesRef;
 import org.apache.solr.common.MapWriter;
 import org.apache.solr.schema.FieldType;
+import org.apache.solr.search.DocValuesIteratorCache;
 
 class BoolFieldWriter extends StringFieldWriter {
-  public BoolFieldWriter(String field, FieldType fieldType) {
-    super(field, fieldType);
+  public BoolFieldWriter(
+      String field,
+      FieldType fieldType,
+      DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache) {
+    super(field, fieldType, docValuesCache);
   }
 
   @Override

solr/core/src/java/org/apache/solr/handler/export/DateFieldWriter.java

@@ -17,58 +17,19 @@
 
 package org.apache.solr.handler.export;
 
-import com.carrotsearch.hppc.IntObjectHashMap;
 import java.io.IOException;
 import java.util.Date;
-import org.apache.lucene.index.DocValues;
-import org.apache.lucene.index.LeafReaderContext;
-import org.apache.lucene.index.NumericDocValues;
 import org.apache.solr.common.MapWriter;
+import org.apache.solr.search.DocValuesIteratorCache;
 
-class DateFieldWriter extends FieldWriter {
-  private String field;
-  private IntObjectHashMap<NumericDocValues> docValuesCache = new IntObjectHashMap<>();
-
-  public DateFieldWriter(String field) {
-    this.field = field;
+class DateFieldWriter extends LongFieldWriter {
+  public DateFieldWriter(
+      String field, DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache) {
+    super(field, docValuesCache);
   }
 
   @Override
-  public boolean write(
-      SortDoc sortDoc, LeafReaderContext readerContext, MapWriter.EntryWriter ew, int fieldIndex)
-      throws IOException {
-    Long val;
-    SortValue sortValue = sortDoc.getSortValue(this.field);
-    if (sortValue != null) {
-      if (sortValue.isPresent()) {
-        val = (long) sortValue.getCurrentValue();
-      } else { // empty-value
-        return false;
-      }
-    } else {
-      // field is not part of 'sort' param, but part of 'fl' param
-      int readerOrd = readerContext.ord;
-      NumericDocValues vals = null;
-      if (docValuesCache.containsKey(readerOrd)) {
-        NumericDocValues numericDocValues = docValuesCache.get(readerOrd);
-        if (numericDocValues.docID() < sortDoc.docId) {
-          // We have not advanced beyond the current docId so we can use this docValues.
-          vals = numericDocValues;
-        }
-      }
-
-      if (vals == null) {
-        vals = DocValues.getNumeric(readerContext.reader(), this.field);
-        docValuesCache.put(readerOrd, vals);
-      }
-
-      if (vals.advance(sortDoc.docId) == sortDoc.docId) {
-        val = vals.longValue();
-      } else {
-        return false;
-      }
-    }
-    ew.put(this.field, new Date(val));
-    return true;
+  protected void doWrite(MapWriter.EntryWriter ew, long val) throws IOException {
+    ew.put(field, new Date(val));
   }
 }

solr/core/src/java/org/apache/solr/handler/export/DoubleFieldWriter.java

@@ -17,57 +17,46 @@
 
 package org.apache.solr.handler.export;
 
-import com.carrotsearch.hppc.IntObjectHashMap;
 import java.io.IOException;
-import org.apache.lucene.index.DocValues;
 import org.apache.lucene.index.LeafReaderContext;
 import org.apache.lucene.index.NumericDocValues;
 import org.apache.solr.common.MapWriter;
+import org.apache.solr.search.DocValuesIteratorCache;
 
 class DoubleFieldWriter extends FieldWriter {
-  private String field;
-  private IntObjectHashMap<NumericDocValues> docValuesCache = new IntObjectHashMap<>();
+  private final String field;
+  private final DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache;
 
-  public DoubleFieldWriter(String field) {
+  public DoubleFieldWriter(
+      String field, DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache) {
     this.field = field;
+    this.docValuesCache = docValuesCache;
   }
 
   @Override
   public boolean write(
       SortDoc sortDoc, LeafReaderContext readerContext, MapWriter.EntryWriter ew, int fieldIndex)
       throws IOException {
+    double val;
     SortValue sortValue = sortDoc.getSortValue(this.field);
     if (sortValue != null) {
       if (sortValue.isPresent()) {
-        double val = (double) sortValue.getCurrentValue();
-        ew.put(this.field, val);
-        return true;
+        val = (double) sortValue.getCurrentValue();
       } else { // empty-value
         return false;
       }
     } else {
       // field is not part of 'sort' param, but part of 'fl' param
-      int readerOrd = readerContext.ord;
-      NumericDocValues vals = null;
-      if (docValuesCache.containsKey(readerOrd)) {
-        NumericDocValues numericDocValues = docValuesCache.get(readerOrd);
-        if (numericDocValues.docID() < sortDoc.docId) {
-          // We have not advanced beyond the current docId so we can use this docValues.
-          vals = numericDocValues;
-        }
-      }
-
-      if (vals == null) {
-        vals = DocValues.getNumeric(readerContext.reader(), this.field);
-        docValuesCache.put(readerOrd, vals);
-      }
-      if (vals.advance(sortDoc.docId) == sortDoc.docId) {
-        long val = vals.longValue();
-        ew.put(this.field, Double.longBitsToDouble(val));
-        return true;
+      NumericDocValues vals =
+          docValuesCache.getNumericDocValues(
+              sortDoc.docId, readerContext.reader(), readerContext.ord);
+      if (vals != null) {
+        val = Double.longBitsToDouble(vals.longValue());
       } else {
         return false;
       }
     }
+    ew.put(this.field, val);
+    return true;
   }
 }

solr/core/src/java/org/apache/solr/handler/export/ExportWriter.java

@@ -74,6 +74,7 @@
 import org.apache.solr.schema.SchemaField;
 import org.apache.solr.schema.SortableTextField;
 import org.apache.solr.schema.StrField;
+import org.apache.solr.search.DocValuesIteratorCache;
 import org.apache.solr.search.SolrIndexSearcher;
 import org.apache.solr.search.SortSpec;
 import org.apache.solr.search.SyntaxError;
@@ -99,6 +100,14 @@ public class ExportWriter implements SolrCore.RawWriter, Closeable {
 
   public static final int DEFAULT_BATCH_SIZE = 30000;
   public static final int DEFAULT_QUEUE_SIZE = 150000;
+  private static final FieldWriter EMPTY_FIELD_WRITER =
+      new FieldWriter() {
+        @Override
+        public boolean write(
+            SortDoc sortDoc, LeafReaderContext readerContext, EntryWriter out, int fieldIndex) {
+          return false;
+        }
+      };
 
   private OutputStreamWriter respWriter;
   final SolrQueryRequest req;
@@ -480,6 +489,7 @@ public FieldWriter[] getFieldWriters(String[] fields, SolrIndexSearcher searcher
       throws IOException {
     IndexSchema schema = searcher.getSchema();
     FieldWriter[] writers = new FieldWriter[fields.length];
+    DocValuesIteratorCache dvIterCache = new DocValuesIteratorCache(searcher, false);
     for (int i = 0; i < fields.length; i++) {
       String field = fields[i];
       SchemaField schemaField = null;
@@ -501,47 +511,51 @@ public FieldWriter[] getFieldWriters(String[] fields, SolrIndexSearcher searcher
             schemaField + " Must have useDocValuesAsStored='true' to be used with export writer");
       }
 
-      if (fieldType instanceof IntValueFieldType) {
+      DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache = dvIterCache.getSupplier(field);
+
+      if (docValuesCache == null) {
+        writers[i] = EMPTY_FIELD_WRITER;
+      } else if (fieldType instanceof IntValueFieldType) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true, docValuesCache);
         } else {
-          writers[i] = new IntFieldWriter(field);
+          writers[i] = new IntFieldWriter(field, docValuesCache);
         }
       } else if (fieldType instanceof LongValueFieldType) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true, docValuesCache);
         } else {
-          writers[i] = new LongFieldWriter(field);
+          writers[i] = new LongFieldWriter(field, docValuesCache);
         }
       } else if (fieldType instanceof FloatValueFieldType) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true, docValuesCache);
         } else {
-          writers[i] = new FloatFieldWriter(field);
+          writers[i] = new FloatFieldWriter(field, docValuesCache);
         }
       } else if (fieldType instanceof DoubleValueFieldType) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true, docValuesCache);
         } else {
-          writers[i] = new DoubleFieldWriter(field);
+          writers[i] = new DoubleFieldWriter(field, docValuesCache);
         }
       } else if (fieldType instanceof StrField || fieldType instanceof SortableTextField) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, false);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, false, docValuesCache);
         } else {
-          writers[i] = new StringFieldWriter(field, fieldType);
+          writers[i] = new StringFieldWriter(field, fieldType, docValuesCache);
         }
       } else if (fieldType instanceof DateValueFieldType) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, false);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, false, docValuesCache);
         } else {
-          writers[i] = new DateFieldWriter(field);
+          writers[i] = new DateFieldWriter(field, docValuesCache);
         }
       } else if (fieldType instanceof BoolField) {
         if (multiValued) {
-          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true);
+          writers[i] = new MultiFieldWriter(field, fieldType, schemaField, true, docValuesCache);
         } else {
-          writers[i] = new BoolFieldWriter(field, fieldType);
+          writers[i] = new BoolFieldWriter(field, fieldType, docValuesCache);
         }
       } else {
         throw new IOException(

solr/core/src/java/org/apache/solr/handler/export/FloatFieldWriter.java

@@ -17,58 +17,46 @@
 
 package org.apache.solr.handler.export;
 
-import com.carrotsearch.hppc.IntObjectHashMap;
 import java.io.IOException;
-import org.apache.lucene.index.DocValues;
 import org.apache.lucene.index.LeafReaderContext;
 import org.apache.lucene.index.NumericDocValues;
 import org.apache.solr.common.MapWriter;
+import org.apache.solr.search.DocValuesIteratorCache;
 
 class FloatFieldWriter extends FieldWriter {
-  private String field;
-  private IntObjectHashMap<NumericDocValues> docValuesCache = new IntObjectHashMap<>();
+  private final String field;
+  private final DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache;
 
-  public FloatFieldWriter(String field) {
+  public FloatFieldWriter(
+      String field, DocValuesIteratorCache.FieldDocValuesSupplier docValuesCache) {
     this.field = field;
+    this.docValuesCache = docValuesCache;
   }
 
   @Override
   public boolean write(
       SortDoc sortDoc, LeafReaderContext readerContext, MapWriter.EntryWriter ew, int fieldIndex)
       throws IOException {
+    float val;
     SortValue sortValue = sortDoc.getSortValue(this.field);
     if (sortValue != null) {
       if (sortValue.isPresent()) {
-        float val = (float) sortValue.getCurrentValue();
-        ew.put(this.field, val);
-        return true;
+        val = (float) sortValue.getCurrentValue();
       } else { // empty-value
         return false;
       }
     } else {
       // field is not part of 'sort' param, but part of 'fl' param
-      int readerOrd = readerContext.ord;
-      NumericDocValues vals = null;
-      if (docValuesCache.containsKey(readerOrd)) {
-        NumericDocValues numericDocValues = docValuesCache.get(readerOrd);
-        if (numericDocValues.docID() < sortDoc.docId) {
-          // We have not advanced beyond the current docId so we can use this docValues.
-          vals = numericDocValues;
-        }
-      }
-
-      if (vals == null) {
-        vals = DocValues.getNumeric(readerContext.reader(), this.field);
-        docValuesCache.put(readerOrd, vals);
-      }
-
-      if (vals.advance(sortDoc.docId) == sortDoc.docId) {
-        int val = (int) vals.longValue();
-        ew.put(this.field, Float.intBitsToFloat(val));
-        return true;
+      NumericDocValues vals =
+          docValuesCache.getNumericDocValues(
+              sortDoc.docId, readerContext.reader(), readerContext.ord);
+      if (vals != null) {
+        val = Float.intBitsToFloat((int) vals.longValue());
       } else {
         return false;
       }
     }
+    ew.put(this.field, val);
+    return true;
   }
 }