I have written a blog on this a while ago where I was writing to multiple tables from my mapper class. You can look into it at http://bigdatabuzz.wordpress.com/2012/04/24/how-to-write-to-multiple-hbase-tables-in-a-mapreduce-job/
Key things are, a) job.setOutputFormatClass (MultiTableOutputFormat.class); b) In your reducer , you would stick with the if else to write to the corresponding HBase tables. Thanks Ravi Magham On Wed, Aug 28, 2013 at 3:50 AM, jamal sasha <[email protected]> wrote: > Hi, > I am new to hbase and am trying to achieve the following. > > I am reading data from hdfs in mapper and parsing it.. > > So, in reducer I want my output to write to hbase instead of hdfs > But here is the thing. > > public static class MyTableReducer extends TableReducer<Text, Text, > ImmutableBytesWritable> { > > public void reduce(Text key, Iterable<Text> values, Context context) > throws IOException, InterruptedException { > int type = getType(values.toString()); > if (type == 1) // put data to table 1 > if (type==2) // put data to table 2 > > > } > } > > How do I do this? > Thanks >
