Mailing-List: contact commits-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hbase.apache.org
Content-Type: text/plain; charset="utf-8"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Subject: svn commit: r997975 - in /hbase/trunk: CHANGES.txt
 src/main/java/org/apache/hadoop/hbase/io/hfile/HFile.java
Date: Fri, 17 Sep 2010 04:21:28 -0000
To: commits@hbase.apache.org
From: stack@apache.org
Message-Id: <20100917042129.07D5023889E2@eris.apache.org>

Author: stack
Date: Fri Sep 17 04:21:28 2010
New Revision: 997975

URL: http://svn.apache.org/viewvc?rev=997975&view=rev
Log:
HBASE-3006 Reading compressed HFile blocks causes way too many DFS RPC calls severly impacting performance--Now add fix I intended, a spelling mistake in HFile

Modified:
    hbase/trunk/CHANGES.txt
    hbase/trunk/src/main/java/org/apache/hadoop/hbase/io/hfile/HFile.java

Modified: hbase/trunk/CHANGES.txt
URL: http://svn.apache.org/viewvc/hbase/trunk/CHANGES.txt?rev=997975&r1=997974&r2=997975&view=diff
==============================================================================
--- hbase/trunk/CHANGES.txt (original)
+++ hbase/trunk/CHANGES.txt Fri Sep 17 04:21:28 2010
@@ -523,6 +523,9 @@ Release 0.21.0 - Unreleased
    HBASE-2986  multi writable can npe causing client hang
    HBASE-2979  Fix failing TestMultParrallel in hudson build
    HBASE-2899  hfile.min.blocksize.size ignored/documentation wrong
+   HBASE-3006  Reading compressed HFile blocks causes way too many DFS RPC
+               calls severly impacting performance
+               (Kannan Muthukkaruppan via Stack)
 
   IMPROVEMENTS
    HBASE-1760  Cleanup TODOs in HTable

Modified: hbase/trunk/src/main/java/org/apache/hadoop/hbase/io/hfile/HFile.java
URL: http://svn.apache.org/viewvc/hbase/trunk/src/main/java/org/apache/hadoop/hbase/io/hfile/HFile.java?rev=997975&r1=997974&r2=997975&view=diff
==============================================================================
--- hbase/trunk/src/main/java/org/apache/hadoop/hbase/io/hfile/HFile.java (original)
+++ hbase/trunk/src/main/java/org/apache/hadoop/hbase/io/hfile/HFile.java Fri Sep 17 04:21:28 2010
@@ -19,6 +19,7 @@
  */
 package org.apache.hadoop.hbase.io.hfile;
 
+import java.io.BufferedInputStream;
 import java.io.Closeable;
 import java.io.DataInputStream;
 import java.io.DataOutputStream;
@@ -1051,10 +1052,15 @@ public class HFile {
         // decompressor reading into next block -- IIRC, it just grabs a
         // bunch of data w/o regard to whether decompressor is coming to end of a
         // decompression.
+
+        // We use a buffer of DEFAULT_BLOCKSIZE size.  This might be extreme.
+        // Could maybe do with less. Study and figure it: TODO
         InputStream is = this.compressAlgo.createDecompressionStream(
-          new BoundedRangeFileInputStream(this.istream, offset, compressedSize,
-            pread),
-          decompressor, 0);
+            new BufferedInputStream(
+                new BoundedRangeFileInputStream(this.istream, offset, compressedSize,
+                                                pread),
+                Math.min(DEFAULT_BLOCKSIZE, compressedSize)),
+            decompressor, 0);
         buf = ByteBuffer.allocate(decompressedSize);
         IOUtils.readFully(is, buf.array(), 0, buf.capacity());
         is.close();