HBase: Bulk upload using ImportTsv in HBase Table

ImportTsv helps you to upload data in HBase from a tsv file placed on hdfs.

Lets take and example hbase table , "employee" with column family "basic_info"

Following is an example file on hdfs:

$ hadoop fs -cat /user/hadoop/importtsv

1    emp1    24

2    emp2    26

3    emp3    24

Uploading data int o employee table using ImportTsv

$HBASE_HOME/bin/hbase org.apache.hadoop.hbase.mapreduce.ImportTsv -Dimporttsv.columns=HBASE_ROW_KEY,basic_info:empname,basic_info:age employee hdfs/namenode:address/user/hadoop/importtsv

First column of the file will be stored as  row key, second column will be stored as  "empname" and third column will be stored as age as follows:

hbase(main):013:0>scan 'employee'

ROW                                                  COLUMN+CELL                                                                                                                                             

 1                                                   column=basic_info:age, timestamp=1360238054613, value=24                                                                                               

 1                                                   column=basic_info:empname, timestamp=1360238054613, value=emp1                                                                                       

 2                                                   column=basic_info:age, timestamp=1360238054613, value=26                                                                                               

 2                                                   column=basic_info:empname, timestamp=1360238054613, value=emp2                                                                                  

 3                                                  column=basic_info:age, timestamp=1360238054613, value=24                                                                                               

 3                                                   column=basic_info:empname, timestamp=1360238054613, value=emp3