Table-to-File Example

This example uses mapreduce to extract specified columns from an existing table.

To run this example you will need some data in a table. The following will put a trivial amount of data into accumulo using the accumulo shell:

$ ./bin/accumulo shell -u username -p password
Shell - Apache Accumulo Interactive Shell
- version: 1.5.0
- instance name: instance
- instance id: 00000000-0000-0000-0000-000000000000
- 
- type 'help' for a list of available commands
- 
username@instance> createtable input
username@instance> insert dog cf cq dogvalue
username@instance> insert cat cf cq catvalue
username@instance> insert junk family qualifier junkvalue
username@instance> quit

The TableToFile class configures a map-only job to read the specified columns and write the key/value pairs to a file in HDFS.

The following will extract the rows containing the column “cf:cq”:

$ bin/tool.sh lib/accumulo-examples-simple.jar org.apache.accumulo.examples.simple.mapreduce.TableToFile -u user -p passwd -i instance -t input --columns cf:cq --output /tmp/output

$ hadoop fs -ls /tmp/output
-rw-r--r--   1 username supergroup          0 2013-01-10 14:44 /tmp/output/_SUCCESS
drwxr-xr-x   - username supergroup          0 2013-01-10 14:44 /tmp/output/_logs
drwxr-xr-x   - username supergroup          0 2013-01-10 14:44 /tmp/output/_logs/history
-rw-r--r--   1 username supergroup       9049 2013-01-10 14:44 /tmp/output/_logs/history/job_201301081658_0011_1357847072863_username_TableToFile%5F1357847071434
-rw-r--r--   1 username supergroup      26172 2013-01-10 14:44 /tmp/output/_logs/history/job_201301081658_0011_conf.xml
-rw-r--r--   1 username supergroup         50 2013-01-10 14:44 /tmp/output/part-m-00000

We can see the output of our little map-reduce job:

$ hadoop fs -text /tmp/output/output/part-m-00000
catrow cf:cq []    catvalue
dogrow cf:cq []    dogvalue
$