Retrieve through a scroll request the whole index student_index
Retrieve through a scroll request the whole index finance_index
make consolidated objects using a loop on student_index data
bulk that new list into a new index.
23000 documents x 3 is easy to process in memory by a program in c# or java..
if there are too many documents, the algo could be like that :
begin to scroll the student_index.
At each scroll, make a multiget or search into finance_data on Roll Number retrieved.
Consolidate
Bulk
Next scroll...
Anyway, it would be interesting to directly build one consolidated index... like indexing a 'final_student_index' the equivalent of student_index, and, instead of mounting the finance_index, do an update of 'final_student_index' to add the "Fee" field.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant logo are trademarks of the Apache Software Foundation in the United States and/or other countries.