[hbase] branch branch-2 updated: HBASE-21810 bulkload support set hfile compression on client

[zg...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

zghao pushed a commit to branch branch-2
in repository https://gitbox.apache.org/repos/asf/hbase.git


The following commit(s) were added to refs/heads/branch-2 by this push:
     new b6afc5f  HBASE-21810 bulkload support set hfile compression on client
b6afc5f is described below

commit b6afc5fbc1e7343192d8eaa355ee21b0b760f722
Author: chenyechao <ch...@gmail.com>
AuthorDate: Mon Mar 11 19:37:17 2019 +0800

    HBASE-21810 bulkload support set hfile compression on client
    
    Signed-off-by: Guanghao Zhang <zg...@apache.org>
---
 .../hadoop/hbase/mapreduce/HFileOutputFormat2.java | 14 ++++++--
 .../hbase/mapreduce/TestHFileOutputFormat2.java    | 42 ++++++++++++++++++++++
 2 files changed, 54 insertions(+), 2 deletions(-)

diff --git a/hbase-mapreduce/src/main/java/org/apache/hadoop/hbase/mapreduce/HFileOutputFormat2.java b/hbase-mapreduce/src/main/java/org/apache/hadoop/hbase/mapreduce/HFileOutputFormat2.java
index cdccb13..59b1994 100644
--- a/hbase-mapreduce/src/main/java/org/apache/hadoop/hbase/mapreduce/HFileOutputFormat2.java
+++ b/hbase-mapreduce/src/main/java/org/apache/hadoop/hbase/mapreduce/HFileOutputFormat2.java
@@ -159,9 +159,11 @@ public class HFileOutputFormat2
   // This constant is public since the client can modify this when setting
   // up their conf object and thus refer to this symbol.
   // It is present for backwards compatibility reasons. Use it only to
-  // override the auto-detection of datablock encoding.
+  // override the auto-detection of datablock encoding and compression.
   public static final String DATABLOCK_ENCODING_OVERRIDE_CONF_KEY =
       "hbase.mapreduce.hfileoutputformat.datablock.encoding";
+  public static final String COMPRESSION_OVERRIDE_CONF_KEY =
+      "hbase.mapreduce.hfileoutputformat.compression";
 
   /**
    * Keep locality while generating HFiles for bulkload. See HBASE-12596
@@ -209,6 +211,13 @@ public class HFileOutputFormat2
         Compression.Algorithm.NONE.getName());
     final Algorithm defaultCompression = HFileWriterImpl
         .compressionByName(defaultCompressionStr);
+    String compressionStr = conf.get(COMPRESSION_OVERRIDE_CONF_KEY);
+    final Algorithm overriddenCompression;
+    if (compressionStr != null) {
+      overriddenCompression = Compression.getCompressionAlgorithmByName(compressionStr);
+    } else {
+      overriddenCompression = null;
+    }
     final boolean compactionExclude = conf.getBoolean(
         "hbase.mapreduce.hfileoutputformat.compaction.exclude", false);
 
@@ -383,7 +392,8 @@ public class HFileOutputFormat2
                   new Path(Bytes.toString(tableName), Bytes.toString(family)));
         }
         WriterLength wl = new WriterLength();
-        Algorithm compression = compressionMap.get(tableAndFamily);
+        Algorithm compression = overriddenCompression;
+        compression = compression == null ? compressionMap.get(tableAndFamily) : compression;
         compression = compression == null ? defaultCompression : compression;
         BloomType bloomType = bloomTypeMap.get(tableAndFamily);
         bloomType = bloomType == null ? BloomType.NONE : bloomType;
diff --git a/hbase-mapreduce/src/test/java/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat2.java b/hbase-mapreduce/src/test/java/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat2.java
index 679b231..76c9359 100644
--- a/hbase-mapreduce/src/test/java/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat2.java
+++ b/hbase-mapreduce/src/test/java/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat2.java
@@ -1493,5 +1493,47 @@ public class TestHFileOutputFormat2  {
 
     return null;
   }
+
+  @Test
+  public void TestConfigureCompression() throws Exception {
+    Configuration conf = new Configuration(this.util.getConfiguration());
+    RecordWriter<ImmutableBytesWritable, Cell> writer = null;
+    TaskAttemptContext context = null;
+    Path dir = util.getDataTestDir("TestConfigureCompression");
+    String hfileoutputformatCompression = "gz";
+
+    try {
+      conf.set(HFileOutputFormat2.OUTPUT_TABLE_NAME_CONF_KEY, TABLE_NAMES[0].getNameAsString());
+      conf.setBoolean(HFileOutputFormat2.LOCALITY_SENSITIVE_CONF_KEY, false);
+
+      conf.set(HFileOutputFormat2.COMPRESSION_OVERRIDE_CONF_KEY, hfileoutputformatCompression);
+
+      Job job = Job.getInstance(conf);
+      FileOutputFormat.setOutputPath(job, dir);
+      context = createTestTaskAttemptContext(job);
+      HFileOutputFormat2 hof = new HFileOutputFormat2();
+      writer = hof.getRecordWriter(context);
+      final byte[] b = Bytes.toBytes("b");
+
+      KeyValue kv = new KeyValue(b, b, b, HConstants.LATEST_TIMESTAMP, b);
+      writer.write(new ImmutableBytesWritable(), kv);
+      writer.close(context);
+      writer = null;
+      FileSystem fs = dir.getFileSystem(conf);
+      RemoteIterator<LocatedFileStatus> iterator = fs.listFiles(dir, true);
+      while (iterator.hasNext()) {
+        LocatedFileStatus keyFileStatus = iterator.next();
+        HFile.Reader reader =
+            HFile.createReader(fs, keyFileStatus.getPath(), new CacheConfig(conf), true, conf);
+        assertEquals(reader.getCompressionAlgorithm().getName(), hfileoutputformatCompression);
+      }
+    } finally {
+      if (writer != null && context != null) {
+        writer.close(context);
+      }
+      dir.getFileSystem(conf).delete(dir, true);
+    }
+
+  }
 }