HBASE-21032 ScanResponses contain only one cell each

Amending-Author: Duo Zhang <zhangduo@apache.org>

hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/StoreScanner.java

@@ -175,16 +175,16 @@ public class StoreScanner extends NonReversedNonLazyKeyValueScanner
     this.oldestUnexpiredTS = scan.isRaw() ? 0L : now - scanInfo.getTtl();
     this.minVersions = scanInfo.getMinVersions();
 
-     // We look up row-column Bloom filters for multi-column queries as part of
-     // the seek operation. However, we also look the row-column Bloom filter
-     // for multi-row (non-"get") scans because this is not done in
-     // StoreFile.passesBloomFilter(Scan, SortedSet<byte[]>).
-     this.useRowColBloom = numColumns > 1 || (!get && numColumns == 1);
-     this.maxRowSize = scanInfo.getTableMaxRowSize();
+    // We look up row-column Bloom filters for multi-column queries as part of
+    // the seek operation. However, we also look the row-column Bloom filter
+    // for multi-row (non-"get") scans because this is not done in
+    // StoreFile.passesBloomFilter(Scan, SortedSet<byte[]>).
+    this.useRowColBloom = numColumns > 1 || (!get && numColumns == 1);
+    this.maxRowSize = scanInfo.getTableMaxRowSize();
     if (get) {
       this.readType = Scan.ReadType.PREAD;
       this.scanUsePread = true;
-    } else if(scanType != ScanType.USER_SCAN) {
+    } else if (scanType != ScanType.USER_SCAN) {
       // For compaction scanners never use Pread as already we have stream based scanners on the
       // store files to be compacted
       this.readType = Scan.ReadType.STREAM;
@@ -556,7 +556,7 @@ public class StoreScanner extends NonReversedNonLazyKeyValueScanner
       // Or if the preadMaxBytes is reached and we may want to return so we can switch to stream in
       // the shipped method below.
       if (kvsScanned % cellsPerHeartbeatCheck == 0 || (scanUsePread &&
-        scan.getReadType() == Scan.ReadType.DEFAULT && bytesRead > preadMaxBytes)) {
+        readType == Scan.ReadType.DEFAULT && bytesRead > preadMaxBytes)) {
         if (scannerContext.checkTimeLimit(LimitScope.BETWEEN_CELLS)) {
           return scannerContext.setScannerState(NextState.TIME_LIMIT_REACHED).hasMoreValues();
         }
@@ -568,7 +568,7 @@ public class StoreScanner extends NonReversedNonLazyKeyValueScanner
       checkScanOrder(prevCell, cell, comparator);
       int cellSize = PrivateCellUtil.estimatedSerializedSizeOf(cell);
       bytesRead += cellSize;
-      if (scanUsePread && scan.getReadType() == Scan.ReadType.DEFAULT &&
+      if (scanUsePread && readType == Scan.ReadType.DEFAULT &&
         bytesRead > preadMaxBytes) {
         // return immediately if we want to switch from pread to stream. We need this because we can
         // only switch in the shipped method, if user use a filter to filter out everything and rpc

hbase-server/src/test/java/org/apache/hadoop/hbase/client/TestGetScanPartialResult.java

@@ -0,0 +1,102 @@
+/**
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.hadoop.hbase.client;
+
+import static org.junit.Assert.assertEquals;
+import static org.junit.Assert.assertTrue;
+
+import java.io.IOException;
+import org.apache.hadoop.hbase.HBaseClassTestRule;
+import org.apache.hadoop.hbase.HBaseTestingUtility;
+import org.apache.hadoop.hbase.TableName;
+import org.apache.hadoop.hbase.testclassification.ClientTests;
+import org.apache.hadoop.hbase.testclassification.MediumTests;
+import org.junit.AfterClass;
+import org.junit.BeforeClass;
+import org.junit.ClassRule;
+import org.junit.Test;
+import org.junit.experimental.categories.Category;
+
+/**
+ * Testcase for HBASE-21032, where use the wrong readType from a Scan instance which is actually a
+ * get scan and cause returning only 1 cell per rpc call.
+ */
+@Category({ ClientTests.class, MediumTests.class })
+public class TestGetScanPartialResult {
+
+  @ClassRule
+  public static final HBaseClassTestRule CLASS_RULE =
+    HBaseClassTestRule.forClass(TestGetScanPartialResult.class);
+
+  private final static HBaseTestingUtility TEST_UTIL = new HBaseTestingUtility();
+  private static final TableName TABLE = TableName.valueOf("table");
+  private static final byte[] CF = { 'c', 'f' };
+  private static final byte[] ROW = { 'r', 'o', 'w' };
+  private static final int VALUE_SIZE = 10000;
+  private static final int NUM_COLUMNS = 300;
+
+  @BeforeClass
+  public static void setUp() throws Exception {
+    TEST_UTIL.startMiniCluster(1);
+    TEST_UTIL.createTable(TABLE, CF);
+  }
+
+  @AfterClass
+  public static void tearDown() throws Exception {
+    TEST_UTIL.shutdownMiniCluster();
+  }
+
+  private static byte[] makeLargeValue(int size) {
+    byte[] v = new byte[size];
+    for (int i = 0; i < size; i++) {
+      v[i] = 0;
+    }
+    return v;
+  }
+
+  @Test
+  public void test() throws IOException {
+    try (Table t = TEST_UTIL.getConnection().getTable(TABLE)) {
+      // populate a row with bunch of columns and large values
+      // to cause scan to return partials
+      byte[] val = makeLargeValue(VALUE_SIZE);
+      Put p = new Put(ROW);
+      for (int i = 0; i < NUM_COLUMNS; i++) {
+        p.addColumn(CF, Integer.toString(i).getBytes(), val);
+      }
+      t.put(p);
+
+      Scan scan = new Scan();
+      scan.withStartRow(ROW);
+      scan.withStopRow(ROW, true);
+      scan.setAllowPartialResults(true);
+      scan.setMaxResultSize(2L * 1024 * 1024);
+      scan.readVersions(1);
+      ResultScanner scanner = t.getScanner(scan);
+
+      int nResults = 0;
+      int nCells = 0;
+      for (Result result = scanner.next(); (result != null); result = scanner.next()) {
+        nResults++;
+        nCells += result.listCells().size();
+      }
+      assertEquals(NUM_COLUMNS, nCells);
+      assertTrue(nResults < 5);
+    }
+  }
+}