View Javadoc

1   /**
2    *
3    * Licensed to the Apache Software Foundation (ASF) under one
4    * or more contributor license agreements.  See the NOTICE file
5    * distributed with this work for additional information
6    * regarding copyright ownership.  The ASF licenses this file
7    * to you under the Apache License, Version 2.0 (the
8    * "License"); you may not use this file except in compliance
9    * with the License.  You may obtain a copy of the License at
10   *
11   *     http://www.apache.org/licenses/LICENSE-2.0
12   *
13   * Unless required by applicable law or agreed to in writing, software
14   * distributed under the License is distributed on an "AS IS" BASIS,
15   * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
16   * See the License for the specific language governing permissions and
17   * limitations under the License.
18   */
19  package org.apache.hadoop.hbase.mapreduce;
20  
21  import java.io.IOException;
22  
23  import org.apache.commons.logging.Log;
24  import org.apache.commons.logging.LogFactory;
25  import org.apache.hadoop.hbase.classification.InterfaceAudience;
26  import org.apache.hadoop.hbase.classification.InterfaceStability;
27  import org.apache.hadoop.conf.Configurable;
28  import org.apache.hadoop.conf.Configuration;
29  import org.apache.hadoop.hbase.HBaseConfiguration;
30  import org.apache.hadoop.hbase.HConstants;
31  import org.apache.hadoop.hbase.TableName;
32  import org.apache.hadoop.hbase.client.Connection;
33  import org.apache.hadoop.hbase.client.ConnectionFactory;
34  import org.apache.hadoop.hbase.client.Delete;
35  import org.apache.hadoop.hbase.client.Mutation;
36  import org.apache.hadoop.hbase.client.Put;
37  import org.apache.hadoop.hbase.client.Table;
38  import org.apache.hadoop.hbase.zookeeper.ZKUtil;
39  import org.apache.hadoop.mapreduce.JobContext;
40  import org.apache.hadoop.mapreduce.OutputCommitter;
41  import org.apache.hadoop.mapreduce.OutputFormat;
42  import org.apache.hadoop.mapreduce.RecordWriter;
43  import org.apache.hadoop.mapreduce.TaskAttemptContext;
44  
45  /**
46   * Convert Map/Reduce output and write it to an HBase table. The KEY is ignored
47   * while the output value <u>must</u> be either a {@link Put} or a
48   * {@link Delete} instance.
49   */
50  @InterfaceAudience.Public
51  @InterfaceStability.Stable
52  public class TableOutputFormat<KEY> extends OutputFormat<KEY, Mutation>
53  implements Configurable {
54  
55    private static final Log LOG = LogFactory.getLog(TableOutputFormat.class);
56  
57    /** Job parameter that specifies the output table. */
58    public static final String OUTPUT_TABLE = "hbase.mapred.outputtable";
59  
60    /**
61     * Optional job parameter to specify a peer cluster.
62     * Used specifying remote cluster when copying between hbase clusters (the
63     * source is picked up from <code>hbase-site.xml</code>).
64     * @see TableMapReduceUtil#initTableReducerJob(String, Class, org.apache.hadoop.mapreduce.Job, Class, String, String, String)
65     */
66    public static final String QUORUM_ADDRESS = "hbase.mapred.output.quorum";
67  
68    /** Optional job parameter to specify peer cluster's ZK client port */
69    public static final String QUORUM_PORT = "hbase.mapred.output.quorum.port";
70  
71    /** Optional specification of the rs class name of the peer cluster */
72    public static final String
73        REGION_SERVER_CLASS = "hbase.mapred.output.rs.class";
74    /** Optional specification of the rs impl name of the peer cluster */
75    public static final String
76        REGION_SERVER_IMPL = "hbase.mapred.output.rs.impl";
77  
78    /** The configuration. */
79    private Configuration conf = null;
80  
81    /**
82     * Writes the reducer output to an HBase table.
83     */
84    protected class TableRecordWriter
85    extends RecordWriter<KEY, Mutation> {
86  
87      private Connection connection;
88      private Table table;
89  
90      /**
91       * @throws IOException 
92       * 
93       */
94      public TableRecordWriter() throws IOException {
95        String tableName = conf.get(OUTPUT_TABLE);
96        this.connection = ConnectionFactory.createConnection(conf);
97        this.table = connection.getTable(TableName.valueOf(tableName));
98        this.table.setAutoFlushTo(false);
99        LOG.info("Created table instance for "  + tableName);
100     }
101     /**
102      * Closes the writer, in this case flush table commits.
103      *
104      * @param context  The context.
105      * @throws IOException When closing the writer fails.
106      * @see org.apache.hadoop.mapreduce.RecordWriter#close(org.apache.hadoop.mapreduce.TaskAttemptContext)
107      */
108     @Override
109     public void close(TaskAttemptContext context)
110     throws IOException {
111       table.close();
112       connection.close();
113     }
114 
115     /**
116      * Writes a key/value pair into the table.
117      *
118      * @param key  The key.
119      * @param value  The value.
120      * @throws IOException When writing fails.
121      * @see org.apache.hadoop.mapreduce.RecordWriter#write(java.lang.Object, java.lang.Object)
122      */
123     @Override
124     public void write(KEY key, Mutation value)
125     throws IOException {
126       if (value instanceof Put) table.put(new Put((Put)value));
127       else if (value instanceof Delete) table.delete(new Delete((Delete)value));
128       else throw new IOException("Pass a Delete or a Put");
129     }
130   }
131 
132   /**
133    * Creates a new record writer.
134    *
135    * @param context  The current task context.
136    * @return The newly created writer instance.
137    * @throws IOException When creating the writer fails.
138    * @throws InterruptedException When the jobs is cancelled.
139    * @see org.apache.hadoop.mapreduce.lib.output.FileOutputFormat#getRecordWriter(org.apache.hadoop.mapreduce.TaskAttemptContext)
140    */
141   @Override
142   public RecordWriter<KEY, Mutation> getRecordWriter(
143     TaskAttemptContext context)
144   throws IOException, InterruptedException {
145     return new TableRecordWriter();
146   }
147 
148   /**
149    * Checks if the output target exists.
150    *
151    * @param context  The current context.
152    * @throws IOException When the check fails.
153    * @throws InterruptedException When the job is aborted.
154    * @see org.apache.hadoop.mapreduce.OutputFormat#checkOutputSpecs(org.apache.hadoop.mapreduce.JobContext)
155    */
156   @Override
157   public void checkOutputSpecs(JobContext context) throws IOException,
158       InterruptedException {
159     // TODO Check if the table exists?
160 
161   }
162 
163   /**
164    * Returns the output committer.
165    *
166    * @param context  The current context.
167    * @return The committer.
168    * @throws IOException When creating the committer fails.
169    * @throws InterruptedException When the job is aborted.
170    * @see org.apache.hadoop.mapreduce.OutputFormat#getOutputCommitter(org.apache.hadoop.mapreduce.TaskAttemptContext)
171    */
172   @Override
173   public OutputCommitter getOutputCommitter(TaskAttemptContext context)
174   throws IOException, InterruptedException {
175     return new TableOutputCommitter();
176   }
177 
178   @Override
179   public Configuration getConf() {
180     return conf;
181   }
182 
183   @Override
184   public void setConf(Configuration otherConf) {
185     this.conf = HBaseConfiguration.create(otherConf);
186 
187     String tableName = this.conf.get(OUTPUT_TABLE);
188     if(tableName == null || tableName.length() <= 0) {
189       throw new IllegalArgumentException("Must specify table name");
190     }
191 
192     String address = this.conf.get(QUORUM_ADDRESS);
193     int zkClientPort = this.conf.getInt(QUORUM_PORT, 0);
194     String serverClass = this.conf.get(REGION_SERVER_CLASS);
195     String serverImpl = this.conf.get(REGION_SERVER_IMPL);
196 
197     try {
198       if (address != null) {
199         ZKUtil.applyClusterKeyToConf(this.conf, address);
200       }
201       if (serverClass != null) {
202         this.conf.set(HConstants.REGION_SERVER_IMPL, serverImpl);
203       }
204       if (zkClientPort != 0) {
205         this.conf.setInt(HConstants.ZOOKEEPER_CLIENT_PORT, zkClientPort);
206       }
207     } catch(IOException e) {
208       LOG.error(e);
209       throw new RuntimeException(e);
210     }
211   }
212 }