Author: jnioche
Date: Thu Nov 14 14:36:12 2013
New Revision: 1541917
URL: http://svn.apache.org/r1541917
Log:
Giving Cleaning and Deduplication jobs a name to display
Modified:
nutch/trunk/src/java/org/apache/nutch/crawl/DeduplicationJob.java
nutch/trunk/src/java/org/apache/nutch/indexer/CleaningJob.java
Modified: nutch/trunk/src/java/org/apache/nutch/crawl/DeduplicationJob.java
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/java/org/apache/nutch/crawl/DeduplicationJob.java?rev=1541917&r1=1541916&r2=1541917&view=diff
==============================================================================
--- nutch/trunk/src/java/org/apache/nutch/crawl/DeduplicationJob.java (original)
+++ nutch/trunk/src/java/org/apache/nutch/crawl/DeduplicationJob.java Thu Nov
14 14:36:12 2013
@@ -228,6 +228,8 @@ public class DeduplicationJob extends Co
+ Integer.toString(new Random().nextInt(Integer.MAX_VALUE)));
JobConf job = new NutchJob(getConf());
+
+ job.setJobName("Deduplication on "+crawldb);
FileInputFormat.addInputPath(job, new Path(crawldb,
CrawlDb.CURRENT_NAME));
Modified: nutch/trunk/src/java/org/apache/nutch/indexer/CleaningJob.java
URL:
http://svn.apache.org/viewvc/nutch/trunk/src/java/org/apache/nutch/indexer/CleaningJob.java?rev=1541917&r1=1541916&r2=1541917&view=diff
==============================================================================
--- nutch/trunk/src/java/org/apache/nutch/indexer/CleaningJob.java (original)
+++ nutch/trunk/src/java/org/apache/nutch/indexer/CleaningJob.java Thu Nov 14
14:36:12 2013
@@ -166,6 +166,8 @@ public class CleaningJob implements Tool
job.setMapOutputValueClass(Text.class);
job.setMapperClass(DBFilter.class);
job.setReducerClass(DeleterReducer.class);
+
+ job.setJobName("CleaningJob");
// need to expicitely allow deletions
job.setBoolean(IndexerMapReduce.INDEXER_DELETE, true);