多输入源(多个scan)配置:在job中配置

TableMapReduceUtil.initTableMapperJob(model.getScans(), model.getMapperClass(), model.getOutputKeyClass(),
				model.getOutputValueClass(), job);

多输出源配置:job中配置

 job.setOutputValueClass(Put.class);
        job.setOutputFormatClass(MultiTableOutputFormat.class);
        job.setReducerClass(InterReducer.class);

redecu过程中根据条件确定要存储的表名 例如:

ImmutableBytesWritable ib = new ImmutableBytesWritable();
                ib.set(表名);
                context.write(ib, put);