[HBASE-13625] Use HDFS for HFileOutputFormat2 partitioner's path

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.1.0, 1.2.0, 2.0.0
• Fix Version/s: 0.98.13, 1.0.2, 1.2.0, 1.1.1, 2.0.0
• Component/s: mapreduce
• Hadoop Flags: Reviewed
• Release Note: Introduces a new config hbase.fs.tmp.dir which is a directory in HDFS (or default file system) to use as a staging directory for HFileOutputFormat2. This is also used as the default for hbase.bulkload.staging.dir
• Assignee: Stephen Yuan Jiang
• Reporter: Stephen Yuan Jiang
• Created: 05/May/15 23:05
• Updated: 31/Aug/15 22:39
• Resolved: 06/May/15 17:11

Description

HBASE-13010 changed hard-coded '/tmp' in HFileOutputFormat2 partitioner's path to 'hadoop.tmp.dir'. This breaks unit test in Windows.

   static void configurePartitioner(Job job, List<ImmutableBytesWritable> splitPoints)
     ...
     // create the partitions file
-    FileSystem fs = FileSystem.get(job.getConfiguration());
-    Path partitionsPath = new Path("/tmp", "partitions_" + UUID.randomUUID());
+    FileSystem fs = FileSystem.get(conf);
+    Path partitionsPath = new Path(conf.get("hadoop.tmp.dir"), "partitions_" + UUID.randomUUID());

Here is the exception from 1 of the UTs when running against Windows (from branch-1.1) - The ':' is an invalid character in windows file path:

java.lang.IllegalArgumentException: Pathname /C:/hbase-server/target/test-data/d25e2228-8959-43ee-b413-4fa69cdb8032/hadoop_tmp/partitions_fb96c0a0-41e6-4964-a391-738cb761ee3e from C:/hbase-server/target/test-data/d25e2228-8959-43ee-b413-4fa69cdb8032/hadoop_tmp/partitions_fb96c0a0-41e6-4964-a391-738cb761ee3e is not a valid DFS filename.
	at org.apache.hadoop.hdfs.DistributedFileSystem.getPathName(DistributedFileSystem.java:197)
	at org.apache.hadoop.hdfs.DistributedFileSystem.access$000(DistributedFileSystem.java:106)
	at org.apache.hadoop.hdfs.DistributedFileSystem$7.doCall(DistributedFileSystem.java:448)
	at org.apache.hadoop.hdfs.DistributedFileSystem$7.doCall(DistributedFileSystem.java:444)
	at org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
	at org.apache.hadoop.hdfs.DistributedFileSystem.create(DistributedFileSystem.java:444)
	at org.apache.hadoop.hdfs.DistributedFileSystem.create(DistributedFileSystem.java:387)
	at org.apache.hadoop.fs.FileSystem.create(FileSystem.java:909)
	at org.apache.hadoop.io.SequenceFile$Writer.<init>(SequenceFile.java:1074)
	at org.apache.hadoop.io.SequenceFile$RecordCompressWriter.<init>(SequenceFile.java:1374)
	at org.apache.hadoop.io.SequenceFile.createWriter(SequenceFile.java:275)
	at org.apache.hadoop.io.SequenceFile.createWriter(SequenceFile.java:297)
	at org.apache.hadoop.hbase.mapreduce.HFileOutputFormat2.writePartitions(HFileOutputFormat2.java:335)
	at org.apache.hadoop.hbase.mapreduce.HFileOutputFormat2.configurePartitioner(HFileOutputFormat2.java:593)
	at org.apache.hadoop.hbase.mapreduce.HFileOutputFormat2.configureIncrementalLoad(HFileOutputFormat2.java:440)
	at org.apache.hadoop.hbase.mapreduce.HFileOutputFormat2.configureIncrementalLoad(HFileOutputFormat2.java:405)
	at org.apache.hadoop.hbase.mapreduce.ImportTsv.createSubmittableJob(ImportTsv.java:539)
	at org.apache.hadoop.hbase.mapreduce.ImportTsv.run(ImportTsv.java:720)
	at org.apache.hadoop.util.ToolRunner.run(ToolRunner.java:70)
	at org.apache.hadoop.hbase.mapreduce.TestImportTsv.doMROnTableTest(TestImportTsv.java:313)
	at org.apache.hadoop.hbase.mapreduce.TestImportTsv.testBulkOutputWithoutAnExistingTable(TestImportTsv.java:168)

The proposed fix is to use a config to point to a hdfs directory.

Issue Links

• is related to: HBASE-13010 HFileOutputFormat2 partitioner's path is hard-coded as '/tmp'