I am using the below code to write to hbase
jsonDStream.foreachRDD(new Function<JavaRDD<String>, Void>() {
#Override
public Void call(JavaRDD<String> rdd) throws Exception {
DataFrame jsonFrame = sqlContext.jsonRDD(rdd);
DataFrame selecteFieldFrame = jsonFrame.select("id_str","created_at","text");
Configuration config = HBaseConfiguration.create();
config.set("hbase.zookeeper.quorum", "d-9543");
config.set("zookeeper.znode.parent","/hbase-unsecure");
config.set("hbase.zookeeper.property.clientPort", "2181");
final JobConf jobConfig=new JobConf(config,SveAsHadoopDataSetExample.class);
jobConfig.setOutputFormat(TableOutputFormat.class);
jobConfig.set(TableOutputFormat.OUTPUT_TABLE,"tableName");
selecteFieldFrame.javaRDD().mapToPair(new PairFunction<Row, ImmutableBytesWritable, Put>() {
#Override
public Tuple2<ImmutableBytesWritable, Put> call(Row row) throws Exception {
// TODO Auto-generated method stub
return convertToPut(row);
}
}).saveAsHadoopDataset(jobConfig);
return null;
}
});
But when i see zkDump in zookeeper the connections keeps on increasing
any suggestion/pointers will be of a great help!
I have the same problem, it is a hbase bug, I fix it:
change org.apache.hadoop.hbase.mapred.TableOutputFormat to org.apache.hadoop.hbase.mapreduce.TableOutputFormat,
and use org.apache.hadoop.mapreduce.Job, not org.apache.hadoop.mapred.JobConf
this is a sample:
import org.apache.hadoop.mapreduce.Job
import org.apache.hadoop.hbase.mapreduce.TableOutputFormat
val conf = HBaseConfiguration.create()
conf.set("hbase.zookeeper.quorum", zk_hosts)
conf.set("hbase.zookeeper.property.clientPort", zk_port)
conf.set(TableOutputFormat.OUTPUT_TABLE, "TABLE_NAME")
val job = Job.getInstance(conf)
job.setOutputFormatClass(classOf[TableOutputFormat[String]])
formatedLines.map{
case (a,b, c) => {
val row = Bytes.toBytes(a)
val put = new Put(row)
put.setDurability(Durability.SKIP_WAL)
put.addColumn(Bytes.toBytes("cf"), Bytes.toBytes("node"), Bytes.toBytes(b))
put.addColumn(Bytes.toBytes("cf"), Bytes.toBytes("topic"), Bytes.toBytes(c))
(new ImmutableBytesWritable(row), put)
}
}.saveAsNewAPIHadoopDataset(job.getConfiguration)
this may help you!
https://github.com/hortonworks-spark/shc/pull/20/commits/2074067c42c5a454fa4cdeec18c462b5367f23b9
Related
Sorry if the question is solved, but I tried to find it and I haven't had success. There are some similar, but I don't found help where I've seen. I have the next problem:
603 [main] WARN b.s.StormSubmitter - Topology submission exception:
Component: [escribirFichero] subscribes from non-existent stream:
[default] of component [buscamosEnKlout]
Exception in thread "main" java.lang.RuntimeException:
InvalidTopologyException(msg:Component:
[escribirFichero] subscribes from non-existent stream:
[default] of component [buscamosEnKlout])
I can't understand why I have this exception. I declare the bolt "buscamosEnKlout" before I use "escribirFichero". Next to my topology I'll put the elemental lines of the bolts. I know the spout is OK,because a trial-and-error approach.
The code of my topology is:
import backtype.storm.Config;
import backtype.storm.LocalCluster;
import backtype.storm.StormSubmitter;
import backtype.storm.stats.RollingWindow;
import backtype.storm.topology.BoltDeclarer;
import backtype.storm.topology.TopologyBuilder;
import bolt.*;
import spout.TwitterSpout;
import twitter4j.FilterQuery;
public class TwitterTopologia {
private static String consumerKey = "xxx1";
private static String consumerSecret = "xxx2";
private static String accessToken = "yyy1";
private static String accessTokenSecret="yyy2";
public static void main(String[] args) throws Exception {
/**************** SETUP ****************/
String remoteClusterTopologyName = null;
if (args!=null) { ... }
TopologyBuilder builder = new TopologyBuilder();
FilterQuery tweetFilterQuery = new FilterQuery();
tweetFilterQuery.track(new String[]{"Vacaciones","Holy Week", "Semana Santa","Holidays","Vacation"});
tweetFilterQuery.language(new String[]{"en","es"});
TwitterSpout spout = new TwitterSpout(consumerKey, consumerSecret, accessToken, accessTokenSecret, tweetFilterQuery);
KloutBuscador buscamosEnKlout = new KloutBuscador();
FileWriterBolt fileWriterBolt = new FileWriterBolt("idUsuarios.txt");
builder.setSpout("spoutLeerTwitter",spout,1);
builder.setBolt("buscamosEnKlout",buscamosEnKlout,1).shuffleGrouping("spoutLeerTwitter");
builder.setBolt("escribirFichero",fileWriterBolt,1).shuffleGrouping("buscamosEnKlout");
Config conf = new Config();
conf.setDebug(true);
if (args != null && args.length > 0) {
conf.setNumWorkers(3);
StormSubmitter.submitTopology(args[0], conf, builder.createTopology());
}
else {
conf.setMaxTaskParallelism(3);
LocalCluster cluster = new LocalCluster();
cluster.submitTopology("twitter-fun", conf, builder.createTopology());
Thread.sleep(460000);
cluster.shutdown();
}
}
}
Bolt "KloutBuscador", alias "buscamosEnKlout", is the next code:
String text = tuple.getStringByField("id");
String cadenaUrl;
cadenaUrl = "http://api.klout.com/v2/identity.json/twitter?screenName=";
cadenaUrl += text.replaceAll("\\[", "").replaceAll("\\]","");
cadenaUrl += "&key=" + kloutKey;
URL url = new URL(cadenaUrl);
HttpURLConnection c = (HttpURLConnection) url.openConnection();
...........c.setRequestMethod("GET");c.setRequestProperty("Content-length", "0");c.setUseCaches(false);c.setAllowUserInteraction(false);c.connect();
int status = c.getResponseCode();
StringBuilder sb = new StringBuilder();
switch (status) {
case 200:
case 201:
BufferedReader br = new BufferedReader(new InputStreamReader(c.getInputStream()));
String line;
while ((line = br.readLine()) != null) sb.append(line + "\n");
br.close();
}
JSONObject jsonResponse = new JSONObject(sb.toString());
//getJSONArray("id");
String results = jsonResponse.toString();
_collector.emit(new Values(text,results));
And the second bolt, fileWriterBolt, alias "escribirFichero", is the next one:
public void prepare(Map map, TopologyContext topologyContext, OutputCollector outputCollector) {
_collector = outputCollector;
try {
writer = new PrintWriter(filename, "UTF-8");...}...}
public void execute(Tuple tuple) {
writer.println((count++)+":::"+tuple.getValues());
//+"+++"+tweet.getUser().getId()+"__FINAL__"+tweet.getUser().getName()
writer.flush();
// Confirm that this tuple has been treated.
//_collector.ack(tuple);
}
If I pass over the bolt of Klous and only write the result of the spout, it works. I don't understand why the Klous's bolt causes this failure
Your buscamosEnKlout bolt needs to declare the format of the tuples it will emit, as well as which streams it will emit to. You most likely haven't implemented declareOutputFields correctly in that bolt. It should contain something like declarer.declare(new Fields("your-text-field", "your-results-field"))
Spring boot new for me as non web project. please guide me how to code Spark streaming in spring boot, i have already work in java-spark project and want to convert in spring boot non web application. any help or suggestion please.
Here is my Spark config
#Bean
public SparkConf sparkConf() {
SparkConf sparkConf = new SparkConf();
sparkConf.set("spark.app.name", "SparkReceiver"); //The name of application. This will appear in the UI and in log data.
//conf.set("spark.ui.port", "7077"); //Port for application's dashboard, which shows memory and workload data.
sparkConf.set("dynamicAllocation.enabled","false"); //Which scales the number of executors registered with this application up and down based on the workload
//conf.set("spark.cassandra.connection.host", "localhost"); //Cassandra Host Adddress/IP
sparkConf.set("spark.serializer","org.apache.spark.serializer.KryoSerializer"); //For serializing objects that will be sent over the network or need to be cached in serialized form.
sparkConf.set("spark.driver.allowMultipleContexts", "true");
sparkConf.setMaster("local[4]");
return sparkConf;
}
#Bean
public JavaSparkContext javaSparkContext() {
return new JavaSparkContext(sparkConf());
}
#Bean
public SparkSession sparkSession() {
return SparkSession
.builder()
.sparkContext(javaSparkContext().sc())
.appName("Java Spark SQL basic example")
.getOrCreate();
}
#Bean
public JavaStreamingContext javaStreamingContext(){
return new JavaStreamingContext(sparkConf(), new Duration(2000));
}
Here is my testing class
#Autowired
private JavaSparkContext sc;
#Autowired
private SparkSession session;
public void testMessage() throws InterruptedException{
JavaStreamingContext jsc = new JavaStreamingContext(sc, new Duration(2000));
Map<String, String> kafkaParams = new HashMap<String, String>();
kafkaParams.put("zookeeper.connect", "localhost:2181"); //Make all kafka data for this cluster appear under a particular path.
kafkaParams.put("group.id", "testgroup"); //String that uniquely identifies the group of consumer processes to which this consumer belongs
kafkaParams.put("metadata.broker.list", "localhost:9092"); //Producer can find a one or more Brokers to determine the Leader for each topic.
kafkaParams.put("serializer.class", "kafka.serializer.StringEncoder"); //Serializer to use when preparing the message for transmission to the Broker.
kafkaParams.put("request.required.acks", "1"); //Producer to require an acknowledgement from the Broker that the message was received.
Set<String> topics = Collections.singleton("16jnfbtopic");
//Create an input DStream for Receiving data from socket
JavaPairInputDStream<String, String> directKafkaStream = KafkaUtils.createDirectStream(jsc,
String.class,
String.class,
StringDecoder.class,
StringDecoder.class,
kafkaParams, topics);
//Create JavaDStream<String>
JavaDStream<String> msgDataStream = directKafkaStream.map(new Function<Tuple2<String, String>, String>() {
#Override
public String call(Tuple2<String, String> tuple2) {
return tuple2._2();
}
});
//Create JavaRDD<Row>
msgDataStream.foreachRDD(new VoidFunction<JavaRDD<String>>() {
#Override
public void call(JavaRDD<String> rdd) {
JavaRDD<Row> rowRDD = rdd.map(new Function<String, Row>() {
#Override
public Row call(String msg) {
Row row = RowFactory.create(msg);
return row;
}
});
//Create Schema
StructType schema = DataTypes.createStructType(new StructField[] {DataTypes.createStructField("Message", DataTypes.StringType, true)});
Dataset<Row> msgDataFrame = session.createDataFrame(rowRDD, schema);
msgDataFrame.show();
}
});
jsc.start();
jsc.awaitTermination();
while run this app i am getting error please Guide me.
Here is my error log
Eclipse Error Log
I have written a mapreduce program that reads the data from hive table using HCATLOG and writes into HBase. This is a map only job with no reducers. I have ran the program from command line and it works as expected(Created a fat jar to avoid Jar issues). I wanted to integrate it oozie (with Help of HUE) . I have two options to run it
Use Mapreduce Action
Use Java Action
Since my Mapreduce program has a driver method that holds the below code
import org.apache.hadoop.conf.*;
import org.apache.hadoop.io.*;
import org.apache.hadoop.mapreduce.*;
import org.apache.hadoop.util.*;
import org.apache.hadoop.fs.*;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hive.hcatalog.data.schema.HCatSchema;
import org.apache.hive.hcatalog.mapreduce.HCatInputFormat;
import org.apache.hive.hcatalog.mapreduce.HCatOutputFormat;
public class HBaseValdiateInsertDriver {
public static void main(String[] args) throws Exception {
String dbName = "Test";
String tableName = "emp";
Configuration conf = new Configuration();
args = new GenericOptionsParser(conf, args).getRemainingArgs();
Job job = new Job(conf, "HBase Get Put Demo");
job.setInputFormatClass(HCatInputFormat.class);
HCatInputFormat.setInput(job, dbName, tableName, null);
job.setJarByClass(HBaseValdiateInsertDriver.class);
job.setMapperClass(HBaseValdiateInsert.class);
job.setMapOutputKeyClass(Text.class);
job.setMapOutputValueClass(Text.class);
job.setNumReduceTasks(0);
FileInputFormat.addInputPath(job, new Path("maprfs:///user/input"));
FileOutputFormat.setOutputPath(job, new Path("maprfs:///user/output"));
job.waitForCompletion(true);
}
}
How do i specify the driver method in oozie, All that i can see is to specify mapper and reducer class.Can someone guide me how do i set the properties ?
Using java action i can specify my driver class as the main class and get this executed , but i face errors like table not found, HCATLOG jars not found etc. I have include hive-site.xml in the workflow(Using Hue) but i feel the system is not able to pick up the properties. Can someone advise me what all do i have to take care of, are there any other configuration properties that i need to include ?
Also the sample program i referred in cloudera website uses
HCatInputFormat.setInput(job, InputJobInfo.create(dbName,
inputTableName, null));
where as i use the below (I dont see a method that accept the above input
HCatInputFormat.setInput(job, dbName, tableName, null);
Below is my mapper code
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.KeyValue;
import org.apache.hadoop.hbase.client.Durability;
import org.apache.hadoop.hbase.client.Get;
import org.apache.hadoop.hbase.client.HTableInterface;
import org.apache.hadoop.hbase.client.Put;
import org.apache.hadoop.hbase.client.Result;
import org.apache.hadoop.hbase.util.Bytes;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.WritableComparable;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hive.hcatalog.data.HCatRecord;
public class HBaseValdiateInsert extends Mapper<WritableComparable, HCatRecord, Text, Text> {
static HTableInterface table;
static HTableInterface inserted;
private String hbaseDate = null;
String existigValue=null;
List<Put> putList = new ArrayList<Put>();
#Override
public void setup(Context context) throws IOException {
Configuration conf = context.getConfiguration();
String tablename = "dev_arch186";
Utils.getHBConnection();
table = Utils.getTable(tablename);
table.setAutoFlushTo(false);
}
#Override
public void cleanup(Context context) {
try {
table.put(putList);
table.flushCommits();
table.close();
} catch (IOException e) {
e.printStackTrace();
}
Utils.closeConnection();
}
#Override
public void map(WritableComparable key, HCatRecord value, Context context) throws IOException, InterruptedException {
String name_hive = (String) value.get(0);
String id_hive = (String) value.get(1);
String rec[] = test.toString().split(",");
Get g = new Get(Bytes.toBytes(name_hive));
existigValue=getOneRecord(Bytes.toBytes("Info"),Bytes.toBytes("name"),name_hive);
if (existigValue.equalsIgnoreCase("NA") || !existigValue.equalsIgnoreCase(id_hive)) {
Put put = new Put(Bytes.toBytes(rec[0]));
put.add(Bytes.toBytes("Info"),
Bytes.toBytes("name"),
Bytes.toBytes(rec[1]));
put.setDurability(Durability.SKIP_WAL);
putList.add(put);
if(putList.size()>25000){
table.put(putList);
table.flushCommits();
}
}
}
public String getOneRecord(byte[] columnFamily, byte[] columnQualifier, String rowKey)
throws IOException {
Get get = new Get(rowKey.getBytes());
get.setMaxVersions(1);
Result rs = table.get(get);
rs.getColumn(columnFamily, columnQualifier);
System.out.println(rs.containsColumn(columnFamily, columnQualifier));
KeyValue result = rs.getColumnLatest(columnFamily,columnQualifier);
if (rs.containsColumn(columnFamily, columnQualifier))
return (Bytes.toString(result.getValue()));
else
return "NA";
}
public boolean columnQualifierExists(String tableName, String ColumnFamily,
String ColumnQualifier, String rowKey) throws IOException {
Get get = new Get(rowKey.getBytes());
Result rs = table.get(get);
return(rs.containsColumn(ColumnFamily.getBytes(),ColumnQualifier.getBytes()));
}
}
Note:
I use MapR (M3) Cluster with HUE as the interface for oozie.
Hive Version : 1-0
HCAT Version: 1-0
I couldn't find any way to initialize HCatInputFormat from Oozie mapreduce action.
But I have a workaround as below.
Created LazyHCatInputFormat by extending HCatInputFormat.
Override the getJobInfo method, to handle initalization. This will be called as part of getSplits(..) call.
private static void lazyInit(Configuration conf){
try{
if(conf==null){
conf = new Configuration(false);
}
conf.addResource(new Path(System.getProperty("oozie.action.conf.xml")));
conf.addResource(new org.apache.hadoop.fs.Path("hive-config.xml"));
String databaseName = conf.get("LazyHCatInputFormat.databaseName");
String tableName = conf.get("LazyHCatInputFormat.tableName");
String partitionFilter = conf.get("LazyHCatInputFormat.partitionFilter");
setInput(conf, databaseName, tableName);
//setFilter(partitionFilter);
//System.out.println("After lazyinit : "+conf.get("mapreduce.lib.hcat.job.info"));
}catch(Exception e){
System.out.println("*** LAZY INIT FAILED ***");
//e.printStackTrace();
}
}
public static InputJobInfo getJobInfo(Configuration conf)
throws IOException {
String jobString = conf.get("mapreduce.lib.hcat.job.info");
if (jobString == null) {
lazyInit(conf);
jobString = conf.get("mapreduce.lib.hcat.job.info");
if(jobString == null){
throw new IOException("job information not found in JobContext. HCatInputFormat.setInput() not called?");
}
}
return (InputJobInfo) HCatUtil.deserialize(jobString);
}
In the oozie map-redcue action, configured as below.
<property>
<name>mapreduce.job.inputformat.class</name>
<value>com.xyz.LazyHCatInputFormat</value>
</property>
<property>
<name>LazyHCatInputFormat.databaseName</name>
<value>HCAT DatabaseNameHere</value>
</property>
<property>
<name>LazyHCatInputFormat.tableName</name>
<value>HCAT TableNameHere</value>
</property>
This might not be the best implementation, but a quick hack to make it work.
If comment builder.setBolt then it runs correctly.Please tell where I am getting wrong
TopologyBuilder builder = new TopologyBuilder();
builder.setSpout("words", new TestWordSpout(), 3);
//Calling setBolt
builder.setBolt("exc", new ExclaimBolt(),3)
.allGrouping("words");
Config conf = new Config();
conf.setDebug(false);
//conf.setNumWorkers(2);
LocalCluster cluster = new LocalCluster();
cluster.submitTopology("test", conf, builder.createTopology());
Had not declared output fields in Spout.
After declaring it works fine.
code example(in u bolt class):
#Override
public void declareOutputFields(OutputFieldsDeclarer declarer) {
declarer.declare(new Fields("msg"));//based on u project
}
While I was going through book Hadoop In Action there was an option which states that rather than adding the small files to distributed cache via program this can be done using the -files generic options.
When I tried this in the setup() of my code I get a FileNotFoundException at fs.open() and it shows me a path which am not sure with.
Question is :
If I use -files generic options by default where in HDFS the file gets copied to ?
The code am trying to execute is below..
public class JoinMapSide2 extends Configured implements Tool{
/* Program : JoinMapSide2.java
Description : Passing the small file via GenericOptionsParser
hadoop jar JoinMapSide2.jar -files orders.txt .........
Input : /data/patent/orders.txt(local file system), /data/patent/customers.txt
Output : /MROut/JoinMapSide2
Date : 23/03/2015
*/
protected static class MapClass extends Mapper <Text,Text,NullWritable,Text>{
// hash table to store the key+value from the distributed file or the background data
private Hashtable <String, String> joinData = new Hashtable <String, String>();
// setup function for filling up the joinData for each each map() call
protected void setup(Context context) throws IOException, InterruptedException {
String line;
String[] tokens;
FileSystem fs;
FSDataInputStream fdis;
LineReader joinReader;
Configuration conf;
Text buffer = new Text();
// get configuration
conf = context.getConfiguration();
// get file system related to the configuration
fs = FileSystem.get(conf);
// get all the local cache files distributed as part of the job
URI[] localFiles = context.getCacheFiles();
System.out.println("Cache File Path:"+localFiles[0].toString());
// check if there are any distributed files
// in our case we are sure we will always one so use that only
if (localFiles.length > 0){
// since the file is now on HDFS FSDataInputStream to read through the file
fdis = fs.open(new Path(localFiles[0].toString()));
joinReader = new LineReader(fdis);
// read local file until EOF
try {
while (joinReader.readLine(buffer) > 0) {
line = buffer.toString();
// apply the split pattern only once
tokens = line.split(",",2);
// add key+value into the Hashtable
joinData.put(tokens[0], tokens[1]);
}
} finally {
joinReader.close();
fdis.close();
}
}
else{
System.err.println("No Cache Files are distributed");
}
}
// map function
protected void map(Text key,Text value, Context context) throws IOException, InterruptedException{
NullWritable kNull = null;
String joinValue = joinData.get(key.toString());
if (joinValue != null){
context.write(kNull, new Text(key.toString() + "," + value.toString() + "," + joinValue));
}
}
}
#Override
public int run(String[] args) throws Exception {
if (args.length < 2){
System.err.println("Usage JoinMapSide -files <smallFile> <inputFile> <outputFile>");
}
Path inFile = new Path(args[0]); // input file(customers.txt)
Path outFile = new Path(args[1]); // output file file
Configuration conf = getConf();
// delimiter for the input file
conf.set("mapreduce.input.keyvaluelinerecordreader.key.value.separator", ",");
Job job = Job.getInstance(conf, "Map Side Join2");
// this is not used as the small file is distributed to all the nodes in the cluster using
// generic options parser
// job.addCacheFile(disFile.toUri());
FileInputFormat.addInputPath(job, inFile);
FileOutputFormat.setOutputPath(job, outFile);
job.setInputFormatClass(KeyValueTextInputFormat.class);
job.setOutputKeyClass(NullWritable.class);
job.setOutputValueClass(Text.class);
job.setJarByClass(JoinMapSide2.class);
job.setMapperClass(MapClass.class);
job.setNumReduceTasks(0);
job.waitForCompletion(true);
return 0;
}
public static void main(String args[]) throws Exception {
int ret = ToolRunner.run(new Configuration(), new JoinMapSide2(), args);
System.exit(ret);
}
This is the below exception I see in the trace
Error: java.io.FileNotFoundException: File does not exist: /tmp/hadoop-yarn/staging/shiva/.staging/job_1427126201553_0003/files/orders.txt#orders.txt
at org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:64)
at org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:54)
at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsUpdateTimes(FSNamesystem.java:1795)
at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsInt(FSNamesystem.java:1738)
at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocations(FSNamesystem.java:1718)
at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocations(FSNamesystem.java:1690)
I start the job like
hadoop jar JoinMapSide2.jar -files orders.txt /data/patent/join/customers.txt /MROut/JoinMapSide2
Any directions would be really helpful. Thanks
First you need to move your orders.txt to hdfs and the you have to use -files
Okay after some searching around I did find out there are 2 errors in my code above.
I should not be using FileDataInputStream to read the distributed file as its local to the node running the mapper I should be using File.
I should not be using URI.toString() instead I should be using the symbolic link added to my file which is just orders.txt
I have corrected code listed below hope it helps.
package org.samples.hina.training;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileReader;
import java.io.IOException;
import java.net.URI;
import java.util.Hashtable;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.conf.Configured;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.NullWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.Tool;
import org.apache.hadoop.util.ToolRunner;
public class JoinMapSide2 extends Configured implements Tool{
/* Program : JoinMapSide2.java
Description : To learn Replicated Join using Distributed Cache via Generic Options -files
Input : file:/patent/join/orders1.txt(distributed to all nodes), /data/patent/customers.txt
Output : /MROut/JoinMapSide2
Date : 24/03/2015
*/
protected static class MapClass extends Mapper <Text,Text,NullWritable,Text>{
// hash table to store the key+value from the distributed file or the background data
private Hashtable <String, String> joinData = new Hashtable <String, String>();
// setup function for filling up the joinData for each each map() call
protected void setup(Context context) throws IOException, InterruptedException {
String line;
String[] tokens;
// get all the cache files set in the configuration set in addCacheFile()
URI[] localFiles = context.getCacheFiles();
System.out.println("File1:"+localFiles[0].toString());
// check if there are any distributed files
// in our case we are sure we will always one so use that only
if (localFiles.length > 0){
// read from LOCAL copy
File localFile1 = new File("./orders1.txt");
// created reader to localFile1
BufferedReader joinReader = new BufferedReader(new FileReader(localFile1));
// read local file until EOF
try {
while ((line = joinReader.readLine()) != null){
// apply the split pattern only once
tokens = line.split(",",2);
// add key+value into the Hashtable
joinData.put(tokens[0], tokens[1]);
}
} finally {
joinReader.close();
}
} else{
System.err.println("Local Cache File does not exist");
}
}
// map function
protected void map(Text key,Text value, Context context) throws IOException, InterruptedException{
NullWritable kNull = null;
String joinValue = joinData.get(key.toString());
if (joinValue != null){
context.write(kNull, new Text(key.toString() + "," + value.toString() + "," + joinValue));
}
}
}
#Override
public int run(String[] args) throws Exception {
if (args.length < 2){
System.err.println("Usage JoinMapSide2 <inputFile> <outputFile>");
}
Path inFile = new Path(args[0]); // input file(customers.txt)
Path outFile = new Path(args[1]); // output file file
Configuration conf = getConf();
// delimiter for the input file
conf.set("mapreduce.input.keyvaluelinerecordreader.key.value.separator", ",");
Job job = Job.getInstance(conf, "Map Side Join2");
// add the files orders1.txt, orders2.txt to distributed cache
// the files added by the Generic Options -files
//job.addCacheFile(disFile1);
FileInputFormat.addInputPath(job, inFile);
FileOutputFormat.setOutputPath(job, outFile);
job.setInputFormatClass(KeyValueTextInputFormat.class);
job.setOutputKeyClass(NullWritable.class);
job.setOutputValueClass(Text.class);
job.setJarByClass(JoinMapSide2.class);
job.setMapperClass(MapClass.class);
job.setNumReduceTasks(0);
job.waitForCompletion(true);
return 0;
}
public static void main(String args[]) throws Exception {
int ret = ToolRunner.run(new Configuration(), new JoinMapSide2(), args);
System.exit(ret);
}
}