Reduce footprint of codec instances further (elastic#114072)

Lots of effectively singleton objects here and fields that can be made static, saves a little more on the per-index overhead and might reveal further simplifications.
pmpailis · Oct 4, 2024 · ddfdd40 · ddfdd40
1 parent 38a0711
commit ddfdd40
Show file tree

Hide file tree

Showing 13 changed files with 34 additions and 34 deletions.
diff --git a/.../src/main/java/org/elasticsearch/benchmark/index/codec/tsdb/internal/DecodeBenchmark.java b/.../src/main/java/org/elasticsearch/benchmark/index/codec/tsdb/internal/DecodeBenchmark.java
@@ -12,6 +12,7 @@
 import org.apache.lucene.store.ByteArrayDataInput;
 import org.apache.lucene.store.ByteArrayDataOutput;
 import org.apache.lucene.store.DataOutput;
+import org.elasticsearch.index.codec.tsdb.DocValuesForUtil;
 import org.openjdk.jmh.infra.Blackhole;
 
 import java.io.IOException;
@@ -43,7 +44,7 @@ public void setupInvocation(int bitsPerValue) {
 
     @Override
     public void benchmark(int bitsPerValue, Blackhole bh) throws IOException {
-        forUtil.decode(bitsPerValue, this.dataInput, this.output);
+        DocValuesForUtil.decode(bitsPerValue, this.dataInput, this.output);
         bh.consume(this.output);
     }
 }
diff --git a/server/src/main/java/org/elasticsearch/index/codec/Elasticsearch814Codec.java b/server/src/main/java/org/elasticsearch/index/codec/Elasticsearch814Codec.java
@@ -67,7 +67,7 @@ public Elasticsearch814Codec() {
      */
     public Elasticsearch814Codec(Zstd814StoredFieldsFormat.Mode mode) {
         super("Elasticsearch814", lucene99Codec);
-        this.storedFieldsFormat = new Zstd814StoredFieldsFormat(mode);
+        this.storedFieldsFormat = mode.getFormat();
     }
 
     @Override

diff --git a/server/src/main/java/org/elasticsearch/index/codec/Elasticsearch816Codec.java b/server/src/main/java/org/elasticsearch/index/codec/Elasticsearch816Codec.java
@@ -28,25 +28,27 @@
  */
 public class Elasticsearch816Codec extends CodecService.DeduplicateFieldInfosCodec {
 
+    private static final Lucene912Codec LUCENE_912_CODEC = new Lucene912Codec();
+    private static final PostingsFormat defaultPostingsFormat = new Lucene912PostingsFormat();
+    private static final DocValuesFormat defaultDVFormat = new Lucene90DocValuesFormat();
+    private static final KnnVectorsFormat defaultKnnVectorsFormat = new Lucene99HnswVectorsFormat();
+
     private final StoredFieldsFormat storedFieldsFormat;
 
-    private final PostingsFormat defaultPostingsFormat;
     private final PostingsFormat postingsFormat = new PerFieldPostingsFormat() {
         @Override
         public PostingsFormat getPostingsFormatForField(String field) {
             return Elasticsearch816Codec.this.getPostingsFormatForField(field);
         }
     };
 
-    private final DocValuesFormat defaultDVFormat;
     private final DocValuesFormat docValuesFormat = new PerFieldDocValuesFormat() {
         @Override
         public DocValuesFormat getDocValuesFormatForField(String field) {
             return Elasticsearch816Codec.this.getDocValuesFormatForField(field);
         }
     };
 
-    private final KnnVectorsFormat defaultKnnVectorsFormat;
     private final KnnVectorsFormat knnVectorsFormat = new PerFieldKnnVectorsFormat() {
         @Override
         public KnnVectorsFormat getKnnVectorsFormatForField(String field) {
@@ -64,11 +66,8 @@ public Elasticsearch816Codec() {
      * worse space-efficiency or vice-versa.
      */
     public Elasticsearch816Codec(Zstd814StoredFieldsFormat.Mode mode) {
-        super("Elasticsearch816", new Lucene912Codec());
-        this.storedFieldsFormat = new Zstd814StoredFieldsFormat(mode);
-        this.defaultPostingsFormat = new Lucene912PostingsFormat();
-        this.defaultDVFormat = new Lucene90DocValuesFormat();
-        this.defaultKnnVectorsFormat = new Lucene99HnswVectorsFormat();
+        super("Elasticsearch816", LUCENE_912_CODEC);
+        this.storedFieldsFormat = mode.getFormat();
     }
 
     @Override

diff --git a/server/src/main/java/org/elasticsearch/index/codec/tsdb/DocValuesForUtil.java b/server/src/main/java/org/elasticsearch/index/codec/tsdb/DocValuesForUtil.java
@@ -21,17 +21,10 @@ public class DocValuesForUtil {
     private static final int BITS_IN_FIVE_BYTES = 5 * Byte.SIZE;
     private static final int BITS_IN_SIX_BYTES = 6 * Byte.SIZE;
     private static final int BITS_IN_SEVEN_BYTES = 7 * Byte.SIZE;
-    private final int blockSize;
-    private final byte[] encoded;
+    private static final int blockSize = ES87TSDBDocValuesFormat.NUMERIC_BLOCK_SIZE;
+    private final byte[] encoded = new byte[1024];
 
-    public DocValuesForUtil() {
-        this(ES87TSDBDocValuesFormat.NUMERIC_BLOCK_SIZE);
-    }
-
-    private DocValuesForUtil(int blockSize) {
-        this.blockSize = blockSize;
-        this.encoded = new byte[1024];
-    }
+    public DocValuesForUtil() {}
 
     public static int roundBits(int bitsPerValue) {
         if (bitsPerValue > 24 && bitsPerValue <= 32) {
@@ -74,7 +67,7 @@ private void encodeFiveSixOrSevenBytesPerValue(long[] in, int bitsPerValue, fina
         out.writeBytes(this.encoded, bytesPerValue * in.length);
     }
 
-    public void decode(int bitsPerValue, final DataInput in, long[] out) throws IOException {
+    public static void decode(int bitsPerValue, final DataInput in, long[] out) throws IOException {
         if (bitsPerValue <= 24) {
             ForUtil.decode(bitsPerValue, in, out);
         } else if (bitsPerValue <= 32) {
@@ -88,7 +81,7 @@ public void decode(int bitsPerValue, final DataInput in, long[] out) throws IOEx
         }
     }
 
-    private void decodeFiveSixOrSevenBytesPerValue(int bitsPerValue, final DataInput in, long[] out) throws IOException {
+    private static void decodeFiveSixOrSevenBytesPerValue(int bitsPerValue, final DataInput in, long[] out) throws IOException {
         // NOTE: we expect multibyte values to be written "least significant byte" first
         int bytesPerValue = bitsPerValue / Byte.SIZE;
         long mask = (1L << bitsPerValue) - 1;

diff --git a/server/src/main/java/org/elasticsearch/index/codec/tsdb/ES87TSDBDocValuesEncoder.java b/server/src/main/java/org/elasticsearch/index/codec/tsdb/ES87TSDBDocValuesEncoder.java
@@ -275,7 +275,7 @@ void decodeOrdinals(DataInput in, long[] out, int bitsPerOrd) throws IOException
             Arrays.fill(out, runLen, out.length, v2);
         } else if (encoding == 2) {
             // bit-packed
-            forUtil.decode(bitsPerOrd, in, out);
+            DocValuesForUtil.decode(bitsPerOrd, in, out);
         } else if (encoding == 3) {
             // cycle encoding
             int cycleLength = (int) v1;
@@ -299,7 +299,7 @@ void decode(DataInput in, long[] out) throws IOException {
         final int bitsPerValue = token >>> 3;
 
         if (bitsPerValue != 0) {
-            forUtil.decode(bitsPerValue, in, out);
+            DocValuesForUtil.decode(bitsPerValue, in, out);
         } else {
             Arrays.fill(out, 0L);
         }

diff --git a/server/src/main/java/org/elasticsearch/index/codec/tsdb/ES87TSDBDocValuesProducer.java b/server/src/main/java/org/elasticsearch/index/codec/tsdb/ES87TSDBDocValuesProducer.java
@@ -355,7 +355,7 @@ public TermsEnum termsEnum() throws IOException {
         }
     }
 
-    private abstract class BaseSortedSetDocValues extends SortedSetDocValues {
+    private abstract static class BaseSortedSetDocValues extends SortedSetDocValues {
 
         final SortedSetEntry entry;
         final IndexInput data;

diff --git a/server/src/main/java/org/elasticsearch/index/codec/vectors/ES813FlatVectorFormat.java b/server/src/main/java/org/elasticsearch/index/codec/vectors/ES813FlatVectorFormat.java
@@ -36,7 +36,7 @@ public class ES813FlatVectorFormat extends KnnVectorsFormat {
 
     static final String NAME = "ES813FlatVectorFormat";
 
-    private final FlatVectorsFormat format = new Lucene99FlatVectorsFormat(DefaultFlatVectorScorer.INSTANCE);
+    private static final FlatVectorsFormat format = new Lucene99FlatVectorsFormat(DefaultFlatVectorScorer.INSTANCE);
 
     /**
      * Sole constructor

diff --git a/...rc/main/java/org/elasticsearch/index/codec/vectors/ES814ScalarQuantizedVectorsFormat.java b/...rc/main/java/org/elasticsearch/index/codec/vectors/ES814ScalarQuantizedVectorsFormat.java
@@ -49,6 +49,10 @@ public class ES814ScalarQuantizedVectorsFormat extends FlatVectorsFormat {
 
     private static final FlatVectorsFormat rawVectorFormat = new Lucene99FlatVectorsFormat(DefaultFlatVectorScorer.INSTANCE);
 
+    static final FlatVectorsScorer flatVectorScorer = new ESFlatVectorsScorer(
+        new ScalarQuantizedVectorScorer(DefaultFlatVectorScorer.INSTANCE)
+    );
+
     /** The minimum confidence interval */
     private static final float MINIMUM_CONFIDENCE_INTERVAL = 0.9f;
 
@@ -60,7 +64,6 @@ public class ES814ScalarQuantizedVectorsFormat extends FlatVectorsFormat {
      * calculated as `1-1/(vector_dimensions + 1)`
      */
     public final Float confidenceInterval;
-    final FlatVectorsScorer flatVectorScorer;
 
     private final byte bits;
     private final boolean compress;
@@ -83,7 +86,6 @@ public ES814ScalarQuantizedVectorsFormat(Float confidenceInterval, int bits, boo
             throw new IllegalArgumentException("bits must be one of: 4, 7, 8; bits=" + bits);
         }
         this.confidenceInterval = confidenceInterval;
-        this.flatVectorScorer = new ESFlatVectorsScorer(new ScalarQuantizedVectorScorer(DefaultFlatVectorScorer.INSTANCE));
         this.bits = (byte) bits;
         this.compress = compress;
     }

diff --git a/server/src/main/java/org/elasticsearch/index/codec/vectors/ES815BitFlatVectorFormat.java b/server/src/main/java/org/elasticsearch/index/codec/vectors/ES815BitFlatVectorFormat.java
@@ -22,7 +22,7 @@ public class ES815BitFlatVectorFormat extends KnnVectorsFormat {
 
     static final String NAME = "ES815BitFlatVectorFormat";
 
-    private final FlatVectorsFormat format = new ES815BitFlatVectorsFormat();
+    private static final FlatVectorsFormat format = new ES815BitFlatVectorsFormat();
 
     /**
      * Sole constructor

diff --git a/server/src/main/java/org/elasticsearch/index/codec/vectors/ES815BitFlatVectorsFormat.java b/server/src/main/java/org/elasticsearch/index/codec/vectors/ES815BitFlatVectorsFormat.java
@@ -27,7 +27,7 @@
 
 class ES815BitFlatVectorsFormat extends FlatVectorsFormat {
 
-    private final FlatVectorsFormat delegate = new Lucene99FlatVectorsFormat(FlatBitVectorScorer.INSTANCE);
+    private static final FlatVectorsFormat delegate = new Lucene99FlatVectorsFormat(FlatBitVectorScorer.INSTANCE);
 
     protected ES815BitFlatVectorsFormat() {
         super("ES815BitFlatVectorsFormat");

diff --git a/server/src/main/java/org/elasticsearch/index/codec/vectors/ES815HnswBitVectorsFormat.java b/server/src/main/java/org/elasticsearch/index/codec/vectors/ES815HnswBitVectorsFormat.java
@@ -30,7 +30,7 @@ public class ES815HnswBitVectorsFormat extends KnnVectorsFormat {
     private final int maxConn;
     private final int beamWidth;
 
-    private final FlatVectorsFormat flatVectorsFormat = new ES815BitFlatVectorsFormat();
+    private static final FlatVectorsFormat flatVectorsFormat = new ES815BitFlatVectorsFormat();
 
     public ES815HnswBitVectorsFormat() {
         this(16, 100);

diff --git a/server/src/main/java/org/elasticsearch/index/codec/zstd/Zstd814StoredFieldsFormat.java b/server/src/main/java/org/elasticsearch/index/codec/zstd/Zstd814StoredFieldsFormat.java
@@ -52,17 +52,23 @@ public enum Mode {
         BEST_COMPRESSION(3, BEST_COMPRESSION_BLOCK_SIZE, 2048);
 
         final int level, blockSizeInBytes, blockDocCount;
+        final Zstd814StoredFieldsFormat format;
 
         Mode(int level, int blockSizeInBytes, int blockDocCount) {
             this.level = level;
             this.blockSizeInBytes = blockSizeInBytes;
             this.blockDocCount = blockDocCount;
+            this.format = new Zstd814StoredFieldsFormat(this);
+        }
+
+        public Zstd814StoredFieldsFormat getFormat() {
+            return format;
         }
     }
 
     private final Mode mode;
 
-    public Zstd814StoredFieldsFormat(Mode mode) {
+    private Zstd814StoredFieldsFormat(Mode mode) {
         super("ZstdStoredFields814", new ZstdCompressionMode(mode.level), mode.blockSizeInBytes, mode.blockDocCount, 10);
         this.mode = mode;
     }

diff --git a/server/src/test/java/org/elasticsearch/index/codec/tsdb/DocValuesForUtilTests.java b/server/src/test/java/org/elasticsearch/index/codec/tsdb/DocValuesForUtilTests.java
@@ -70,11 +70,10 @@ public void testEncodeDecode() throws IOException {
         {
             // decode
             IndexInput in = d.openInput("test.bin", IOContext.READONCE);
-            final DocValuesForUtil forUtil = new DocValuesForUtil();
             final long[] restored = new long[ES87TSDBDocValuesFormat.NUMERIC_BLOCK_SIZE];
             for (int i = 0; i < iterations; ++i) {
                 final int bitsPerValue = in.readByte();
-                forUtil.decode(bitsPerValue, in, restored);
+                DocValuesForUtil.decode(bitsPerValue, in, restored);
                 assertArrayEquals(
                     Arrays.toString(restored),
                     ArrayUtil.copyOfSubArray(