MRIT is not designed for that scenario, so you simply can't.
What people usually do is have a process whereby, after
the initial bulk load, there is some way their system-of-record
"knows" what new docs have been added since and
indexes only those. Flume is sometimes used if you have
access.
Best
Hi
I would like to index only new data but not already indexed data(delta
Indexing). how can i achieve it using MRIT.
Thanks in advance
--
View this message in context:
http://lucene.472066.n3.nabble.com/MapReduceIndexerTool-Indexing-tp4248387p4248573.html
Sent from the Solr - User mailing li
Yes it does. MRIT is intended for initial bulk loads. It takes whatever
it's pointed at and indexes it.
Additionally, it does not update documents. If the same document (by
ID) is indexed twice, you'll wind up with two copies in your results.
Best,
Erick
On Mon, Jan 4, 2016 at 5:00 AM, vidya wr
