spark-lucenerdd
spark-lucenerdd copied to clipboard
[lucenerdd] save index to hdfs
WIP
Tries to fix #147
@zouzias Do you need any support with this?
Hi @yeikel , if you want to work on the issue feel free to continue the PR.
The missing items are:
- From every executor copy the index folder to HDFS, i.e.,
/path_in_hdfs/partition_id/lucene_index
. - Implement a LuceneRDD constructor that reads the precomputed index from HDFS.
Any chance you can resolve the merge conflicts here?
Any chance you can resolve the merge conflicts here?
Done