Apache
/
hbase
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HBASE-1901 General partitioner for hbase-48 bulk (behind the api, write hfiles direct) uploader 

git-svn-id: https://svn.apache.org/repos/asf/hadoop/hbase/trunk@883833 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Michael Stack 2009-11-24 19:44:32 +00:00
parent 866453467e
commit 7486a7efe2
5 changed files with 192 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
CHANGES.txt
View File

@ -199,6 +199,10 @@ Release 0.21.0 - Unreleased
HBASE-1961 HBase EC2 scripts
HBASE-1971 Unit test the full WAL replay cycle
NEW FEATURES
HBASE-1901 "General" partitioner for "hbase-48" bulk (behind the api, write
hfiles direct) uploader
OPTIMIZATIONS
HBASE-410 [testing] Speed up the test suite

96
src/java/org/apache/hadoop/hbase/mapreduce/SimpleTotalOrderPartitioner.java Normal file
View File

@ -0,0 +1,96 @@
/**
* Copyright 2009 The Apache Software Foundation
*
* 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.mapreduce;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.conf.Configurable;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.io.ImmutableBytesWritable;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hadoop.mapreduce.Partitioner;
/**
* A partitioner that takes start and end keys and uses bigdecimal to figure
* which reduce a key belongs to. Pass the start and end
* keys in the Configuration using <code>hbase.simpletotalorder.start</code>
* and <code>hbase.simpletotalorder.end</code>. The end key needs to be
* exclusive; i.e. one larger than the biggest key in your key space.
* You may be surprised at how this class partitions the space; it may not
* align with preconceptions; e.g. a start key of zero and an end key of 100
* divided in ten will not make regions whose range is 0-10, 10-20, and so on.
* Make your own partitioner if you need the region spacing to come out a
* particular way.
* @param <VALUE>
* @see #START
* @see #END
*/
public class SimpleTotalOrderPartitioner<VALUE> extends Partitioner<ImmutableBytesWritable, VALUE>
implements Configurable {
private final Log LOG = LogFactory.getLog(this.getClass());
public static final String START = "hbase.simpletotalorder.start";
public static final String END = "hbase.simpletotalorder.end";
private Configuration c;
private byte [] startkey;
private byte [] endkey;
private byte [][] splits;
private int lastReduces = -1;
@Override
public int getPartition(final ImmutableBytesWritable key, final VALUE value,
final int reduces) {
if (reduces == 1) return 0;
if (this.lastReduces != reduces) {
this.splits = Bytes.split(this.startkey, this.endkey, reduces - 1);
for (int i = 0; i < splits.length; i++) {
LOG.info(Bytes.toString(splits[i]));
}
}
int pos = Bytes.binarySearch(this.splits, key.get(), key.getOffset(),
key.getLength(), Bytes.BYTES_RAWCOMPARATOR);
// Below code is from hfile index search.
if (pos < 0) {
pos++;
pos *= -1;
if (pos == 0) {
// falls before the beginning of the file.
throw new RuntimeException("Key outside start/stop range: " +
key.toString());
}
pos--;
}
return pos;
}
@Override
public Configuration getConf() {
return this.c;
}
@Override
public void setConf(Configuration conf) {
this.c = conf;
String startStr = this.c.get(START);
String endStr = this.c.get(END);
LOG.info("startkey=" + startStr + ", endkey=" + endStr);
this.startkey = Bytes.toBytes(startStr);
this.endkey = Bytes.toBytes(endStr);
}
}

20
src/java/org/apache/hadoop/hbase/mapreduce/package-info.java
View File

@ -109,8 +109,8 @@ partitioner.
<h2><a name="bulk">Bulk import writing HFiles directly</a></h2>
<p>If importing into a new table, its possible to by-pass the HBase API
and write your content directly to the filesystem properly formatted as
HBase data files (HFiles). Your import will run faster, perhaps as much
as an order of magnitude faster if not more.
HBase data files (HFiles). Your import will run faster, perhaps an order of
magnitude faster if not more.
</p>
<p>You will need to write a MapReduce job. The map task will know how to
pull from your data source. Your reduce task will need to be hooked up to
@ -131,14 +131,20 @@ BUT the sort is scoped to the particular reducer. Its not a global sort.
Given the default hash Partitioner, if the keys were 0-4 (inclusive), and you
had configured two reducers, reducer 0 would have get keys 0, 2 and 4 whereas
reducer 1 would get keys 1 and 3 (in order). For your bulk import to work,
the keys need to be orderd so reducer 0 gets keys 0-2 and reducer 1 gets keys
3-4 (See TotalOrderPartitioner up in hadoop for more on what this means).
the keys need to be ordered so reducer 0 gets keys 0-2 and reducer 1 gets keys
3-4 (See TotalOrderPartitioner up in hadoop for more on what this means. See
how it runs a sampler step first. You may need to write one of these).
To achieve total ordering, you will likely need to write a Partitioner
that is intimate with your tables key namespace and that knows how
to distribute keys among the reducers so a total order is maintained.
to distribute keys among the reducers so a total order is maintained. If your
keys are distributed with some regularity across a defined key space -- i.e.
you know the start and end keys -- then the {@link SimpleTotalOrderPartitioner}
may be all you need.
</p>
<p>See org.apache.hadoop.hbase.mapreduce.TestHFileOutputFormat for an example that puts together
{@link org.apache.hadoop.hbase.mapreduce.KeyValueSortReducer} and {@link org.apache.hadoop.hbase.mapreduce.HFileOutputFormat}.</p>
<p>See org.apache.hadoop.hbase.mapreduce.TestHFileOutputFormat for an example
that puts together {@link org.apache.hadoop.hbase.mapreduce.KeyValueSortReducer},
{@link SimpleTotalOrderPartitioner}, and
{@link org.apache.hadoop.hbase.mapreduce.HFileOutputFormat}.</p>
<p>HFileOutputFormat writes HFiles. When your MapReduce file finishes, in your
output directory you will have many HFiles. Run the script <code>bin/loadtable.rb</code>

10
src/test/org/apache/hadoop/hbase/mapreduce/TestHFileOutputFormat.java
View File

@ -31,6 +31,7 @@ import org.apache.hadoop.hbase.HBaseTestCase;
import org.apache.hadoop.hbase.KeyValue;
import org.apache.hadoop.hbase.PerformanceEvaluation;
import org.apache.hadoop.hbase.io.ImmutableBytesWritable;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hadoop.io.Writable;
import org.apache.hadoop.mapreduce.InputFormat;
import org.apache.hadoop.mapreduce.InputSplit;
@ -172,6 +173,15 @@ public class TestHFileOutputFormat extends HBaseTestCase {
job.setMapperClass(TestHFileOutputFormat.PEtoKVMapper.class);
job.setMapOutputKeyClass(ImmutableBytesWritable.class);
job.setMapOutputValueClass(KeyValue.class);
// This partitioner doesn't work well for number keys but using it anyways
// just to demonstrate how to configure it.
job.setPartitionerClass(SimpleTotalOrderPartitioner.class);
// Set start and end rows for partitioner.
job.getConfiguration().set(SimpleTotalOrderPartitioner.START,
Bytes.toString(PerformanceEvaluation.format(0)));
int rows = this.conf.getInt("mapred.map.tasks", 1) * ROWSPERSPLIT;
job.getConfiguration().set(SimpleTotalOrderPartitioner.END,
Bytes.toString(PerformanceEvaluation.format(rows)));
job.setReducerClass(KeyValueSortReducer.class);
job.setOutputFormatClass(HFileOutputFormat.class);
FileOutputFormat.setOutputPath(job, this.testDir);

68
src/test/org/apache/hadoop/hbase/mapreduce/TestSimpleTotalOrderPartitioner.java Normal file
View File

@ -0,0 +1,68 @@
/**
* Copyright 2009 The Apache Software Foundation
*
* 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.mapreduce;
import org.apache.hadoop.hbase.HBaseTestCase;
import org.apache.hadoop.hbase.HConstants;
import org.apache.hadoop.hbase.io.ImmutableBytesWritable;
import org.apache.hadoop.hbase.util.Bytes;
/**
* Test of simple partitioner.
*/
public class TestSimpleTotalOrderPartitioner extends HBaseTestCase {
public void testSplit() throws Exception {
String start = "a";
String end = "{";
SimpleTotalOrderPartitioner<byte []> p =
new SimpleTotalOrderPartitioner<byte []>();
this.conf.set(SimpleTotalOrderPartitioner.START, start);
this.conf.set(SimpleTotalOrderPartitioner.END, end);
p.setConf(this.conf);
ImmutableBytesWritable c = new ImmutableBytesWritable(Bytes.toBytes("c"));
// If one reduce, partition should be 0.
int partition = p.getPartition(c, HConstants.EMPTY_BYTE_ARRAY, 1);
assertEquals(0, partition);
// If two reduces, partition should be 0.
partition = p.getPartition(c, HConstants.EMPTY_BYTE_ARRAY, 2);
assertEquals(0, partition);
// Divide in 3.
partition = p.getPartition(c, HConstants.EMPTY_BYTE_ARRAY, 3);
assertEquals(0, partition);
ImmutableBytesWritable q = new ImmutableBytesWritable(Bytes.toBytes("q"));
partition = p.getPartition(q, HConstants.EMPTY_BYTE_ARRAY, 2);
assertEquals(1, partition);
partition = p.getPartition(q, HConstants.EMPTY_BYTE_ARRAY, 3);
assertEquals(2, partition);
// What about end and start keys.
ImmutableBytesWritable startBytes =
new ImmutableBytesWritable(Bytes.toBytes(start));
partition = p.getPartition(startBytes, HConstants.EMPTY_BYTE_ARRAY, 2);
assertEquals(0, partition);
partition = p.getPartition(startBytes, HConstants.EMPTY_BYTE_ARRAY, 3);
assertEquals(0, partition);
ImmutableBytesWritable endBytes =
new ImmutableBytesWritable(Bytes.toBytes("z"));
partition = p.getPartition(endBytes, HConstants.EMPTY_BYTE_ARRAY, 2);
assertEquals(1, partition);
partition = p.getPartition(endBytes, HConstants.EMPTY_BYTE_ARRAY, 3);
assertEquals(2, partition);
}
}