From fee33057f07874a3cee259541f93bc8a6c3f571a Mon Sep 17 00:00:00 2001
From: Kai Xie <gigikie@gmail.com>
Date: Tue, 8 Jan 2019 13:35:17 +0000
Subject: [PATCH] HADOOP-16018. DistCp won't reassemble chunks when blocks per
 chunk > 0.

Contributed by Kai Xie.

(cherry picked from commit 188bebbe7eaec71433548dc47f5a9a995cbea15a)
---
 .../apache/hadoop/tools/DistCpConstants.java  |  4 +++
 .../hadoop/tools/DistCpOptionSwitch.java      |  2 +-
 .../hadoop/tools/TestDistCpOptions.java       | 29 +++++++++++++++++++
 3 files changed, 34 insertions(+), 1 deletion(-)

diff --git a/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpConstants.java b/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpConstants.java
index 212256ccfd7..494609144b2 100644
--- a/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpConstants.java
+++ b/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpConstants.java
@@ -123,6 +123,10 @@ public final class DistCpConstants {
   public static final String CONF_LABEL_COPY_BUFFER_SIZE =
       "distcp.copy.buffer.size";
 
+  /** DistCp Blocks Per Chunk: {@value}. */
+  public static final String CONF_LABEL_BLOCKS_PER_CHUNK =
+      "distcp.blocks.per.chunk";
+
   /**
    * Constants for DistCp return code to shell / consumer of ToolRunner's run
    */
diff --git a/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpOptionSwitch.java b/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpOptionSwitch.java
index 3ce12b264d3..e57e413de33 100644
--- a/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpOptionSwitch.java
+++ b/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCpOptionSwitch.java
@@ -192,7 +192,7 @@ public enum DistCpOptionSwitch {
       new Option("sizelimit", true, "(Deprecated!) Limit number of files " +
               "copied to <= n bytes")),
 
-  BLOCKS_PER_CHUNK("",
+  BLOCKS_PER_CHUNK(DistCpConstants.CONF_LABEL_BLOCKS_PER_CHUNK,
       new Option("blocksperchunk", true, "If set to a positive value, files"
           + "with more blocks than this value will be split into chunks of "
           + "<blocksperchunk> blocks to be transferred in parallel, and "
diff --git a/hadoop-tools/hadoop-distcp/src/test/java/org/apache/hadoop/tools/TestDistCpOptions.java b/hadoop-tools/hadoop-distcp/src/test/java/org/apache/hadoop/tools/TestDistCpOptions.java
index dd8ec697b65..62a2e6d7514 100644
--- a/hadoop-tools/hadoop-distcp/src/test/java/org/apache/hadoop/tools/TestDistCpOptions.java
+++ b/hadoop-tools/hadoop-distcp/src/test/java/org/apache/hadoop/tools/TestDistCpOptions.java
@@ -20,6 +20,7 @@ package org.apache.hadoop.tools;
 
 import java.util.Collections;
 
+import org.apache.hadoop.conf.Configuration;
 import org.junit.Assert;
 import org.junit.Test;
 
@@ -533,4 +534,32 @@ public class TestDistCpOptions {
     builder.withLogPath(logPath).withVerboseLog(true);
     Assert.assertTrue(builder.build().shouldVerboseLog());
   }
+
+  @Test
+  public void testAppendToConf() {
+    final int expectedBlocksPerChunk = 999;
+    final String expectedValForEmptyConfigKey = "VALUE_OF_EMPTY_CONFIG_KEY";
+
+    DistCpOptions options = new DistCpOptions.Builder(
+        Collections.singletonList(
+            new Path("hdfs://localhost:8020/source")),
+        new Path("hdfs://localhost:8020/target/"))
+        .withBlocksPerChunk(expectedBlocksPerChunk)
+        .build();
+
+    Configuration config = new Configuration();
+    config.set("", expectedValForEmptyConfigKey);
+
+    options.appendToConf(config);
+    Assert.assertEquals(expectedBlocksPerChunk,
+        config.getInt(
+            DistCpOptionSwitch
+                .BLOCKS_PER_CHUNK
+                .getConfigLabel(), 0));
+    Assert.assertEquals(
+        "Some DistCpOptionSwitch's config label is empty! " +
+            "Pls ensure the config label is provided when apply to config, " +
+            "otherwise it may not be fetched properly",
+            expectedValForEmptyConfigKey, config.get(""));
+  }
 }