The example suggests that blog posts also would be routed the same way,
therefore the post and comments end up on the same shard.
On which particular shard both the post and comments end up could be
controlled by features described on the allocation page linked above.
Other examples on the ES site have all tweets for the same user routed
using the same value (the example just use the user login id), so
all the tweets end up together.
That's the long way to say that routing can be any value, so you can do
anything you want on a per document basis.
-Paul
On 7/31/2012 10:26 AM, ajasuja wrote:
Hello Everyone,
One of my usecase is to support mapping at the document level i.e.
each object to be indexed
has a mapping associated with it.
Is it possible to have mapping at the per document basis? If yes, how
can we do it using java api
at runtime?
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.