forked from D-Net/dnet-hadoop
simplified import procedure, maps the infospace as hive tables
This commit is contained in:
parent
1e7a2ac41d
commit
a52d5bde4f
|
|
@ -32,11 +32,6 @@
|
|||
<version>1.0.4-SNAPSHOT</version>
|
||||
</dependency>
|
||||
|
||||
<dependency>
|
||||
<groupId>de.javakaffee</groupId>
|
||||
<artifactId>kryo-serializers</artifactId>
|
||||
</dependency>
|
||||
|
||||
</dependencies>
|
||||
|
||||
|
||||
|
|
|
|||
|
|
@ -3,11 +3,11 @@ package eu.dnetlib.dhp.graph;
|
|||
import com.fasterxml.jackson.databind.ObjectMapper;
|
||||
import eu.dnetlib.dhp.application.ArgumentApplicationParser;
|
||||
import org.apache.commons.io.IOUtils;
|
||||
import org.apache.commons.lang.StringUtils;
|
||||
import org.apache.hadoop.io.Text;
|
||||
import org.apache.spark.api.java.JavaRDD;
|
||||
import org.apache.spark.api.java.JavaSparkContext;
|
||||
import org.apache.spark.sql.Encoders;
|
||||
import org.apache.spark.sql.SaveMode;
|
||||
import org.apache.spark.sql.SparkSession;
|
||||
import scala.Tuple2;
|
||||
|
||||
|
|
@ -19,29 +19,27 @@ public class SparkGraphImporterJob {
|
|||
parser.parseArgument(args);
|
||||
final SparkSession spark = SparkSession
|
||||
.builder()
|
||||
.appName("ImportGraph")
|
||||
.appName(SparkGraphImporterJob.class.getSimpleName())
|
||||
.master(parser.get("master"))
|
||||
.config("hive.metastore.uris", parser.get("hive.metastore.uris"))
|
||||
.enableHiveSupport()
|
||||
.getOrCreate();
|
||||
|
||||
final JavaSparkContext sc = new JavaSparkContext(spark.sparkContext());
|
||||
final String inputPath = parser.get("sourcePath");
|
||||
final String outputPath = parser.get("targetPath");
|
||||
|
||||
final String filter = parser.get("filter");
|
||||
|
||||
// Read the input file and convert it into RDD of serializable object
|
||||
final JavaRDD<Tuple2<String, String>> inputRDD = sc.sequenceFile(inputPath, Text.class, Text.class)
|
||||
.map(item -> new Tuple2<>(item._1.toString(), item._2.toString()));
|
||||
|
||||
GraphMappingUtils.types.forEach((name, clazz) -> {
|
||||
if (StringUtils.isNotBlank(filter) || filter.toLowerCase().contains(name)) {
|
||||
final JavaRDD<Tuple2<String, String>> inputRDD = sc.sequenceFile(inputPath + "/" + name, Text.class, Text.class)
|
||||
.map(item -> new Tuple2<>(item._1.toString(), item._2.toString()));
|
||||
spark.createDataset(inputRDD
|
||||
.filter(s -> s._1().equals(clazz.getName()))
|
||||
.map(Tuple2::_2)
|
||||
.map(s -> new ObjectMapper().readValue(s, clazz))
|
||||
.rdd(), Encoders.bean(clazz))
|
||||
.write()
|
||||
.save(outputPath + "/" + name);
|
||||
}
|
||||
.mode(SaveMode.Overwrite)
|
||||
.saveAsTable("openaire." + name);
|
||||
});
|
||||
|
||||
}
|
||||
|
|
|
|||
|
|
@ -1,6 +1,5 @@
|
|||
[
|
||||
{"paramName":"mt", "paramLongName":"master", "paramDescription": "should be local or yarn", "paramRequired": true},
|
||||
{"paramName":"s", "paramLongName":"sourcePath", "paramDescription": "the path of the sequencial file to read", "paramRequired": true},
|
||||
{"paramName":"f", "paramLongName":"filter", "paramDescription": "csv of typology of dataframe to be generated", "paramRequired": false},
|
||||
{"paramName":"t", "paramLongName":"targetPath", "paramDescription": "the path where store DataFrames on HDFS", "paramRequired": true}
|
||||
{"paramName":"h", "paramLongName":"hive.metastore.uris","paramDescription": "the hive metastore uris", "paramRequired": true}
|
||||
]
|
||||
|
|
@ -16,7 +16,7 @@
|
|||
<value>spark2</value>
|
||||
</property>
|
||||
<property>
|
||||
<name>filter</name>
|
||||
<value></value>
|
||||
<name>hive.metastore.uris</name>
|
||||
<value>thrift://iis-cdh5-test-m3.ocean.icm.edu.pl:9083</value>
|
||||
</property>
|
||||
</configuration>
|
||||
|
|
@ -4,10 +4,6 @@
|
|||
<name>sourcePath</name>
|
||||
<description>the source path</description>
|
||||
</property>
|
||||
<property>
|
||||
<name>targetPath</name>
|
||||
<description>the target path</description>
|
||||
</property>
|
||||
<property>
|
||||
<name>sparkDriverMemory</name>
|
||||
<description>memory for driver process</description>
|
||||
|
|
@ -20,10 +16,6 @@
|
|||
<name>sparkExecutorCores</name>
|
||||
<description>number of cores used by single executor</description>
|
||||
</property>
|
||||
<property>
|
||||
<name>filter</name>
|
||||
<description>csv string to filter the entities to generate</description>
|
||||
</property>
|
||||
</parameters>
|
||||
|
||||
<start to="MapGraphIntoDataFrame"/>
|
||||
|
|
@ -41,12 +33,10 @@
|
|||
<name>MapGraphIntoDataFrame</name>
|
||||
<class>eu.dnetlib.dhp.graph.SparkGraphImporterJob</class>
|
||||
<jar>dhp-graph-mapper-${projectVersion}.jar</jar>
|
||||
<spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory=${sparkDriverMemory} --conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener" --conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener"</spark-opts>
|
||||
<spark-opts>--executor-memory ${sparkExecutorMemory} --executor-cores ${sparkExecutorCores} --driver-memory=${sparkDriverMemory} --conf spark.extraListeners="com.cloudera.spark.lineage.NavigatorAppListener" --conf spark.sql.queryExecutionListeners="com.cloudera.spark.lineage.NavigatorQueryListener" --conf spark.sql.warehouse.dir="/user/hive/warehouse"</spark-opts>
|
||||
<arg>-mt</arg> <arg>yarn-cluster</arg>
|
||||
<arg>--sourcePath</arg><arg>${sourcePath}</arg>
|
||||
<arg>--targetPath</arg><arg>${targetPath}</arg>
|
||||
<arg>--filter</arg><arg>${filter}</arg>
|
||||
|
||||
<arg>--hive.metastore.uris</arg><arg>${hive.metastore.uris}</arg>
|
||||
</spark>
|
||||
<ok to="End"/>
|
||||
<error to="Kill"/>
|
||||
|
|
|
|||
Loading…
Reference in New Issue