I've been planning on using ES in a search application, and it could grow
very quickly, so I've been looking at ways to pre-shard my data.
Now, it turns out that every document that I'm indexing has a list of
string facets, and all of my searches will have at least one required
facet, so a natural sharding method would be to shard based on these facets
(if a document has 7 facets, I would assume that it could be available on 7
shards).
Questions:
Is there a way to automatically shard based on facets?
If so, how? If not, could I write a plugin to do so?
If I do shard on facets, will routing to the correct shards be done
automatically, without going to the entire cluster?
If not, is there a way for me to write a plugin to make this happen?
On Wednesday, February 6, 2013 12:59:42 PM UTC-7, Josiah Carlson wrote:
Hello,
I've been planning on using ES in a search application, and it could grow
very quickly, so I've been looking at ways to pre-shard my data.
Now, it turns out that every document that I'm indexing has a list of
string facets, and all of my searches will have at least one required
facet, so a natural sharding method would be to shard based on these facets
(if a document has 7 facets, I would assume that it could be available on 7
shards).
Questions:
Is there a way to automatically shard based on facets?
If so, how? If not, could I write a plugin to do so?
If I do shard on facets, will routing to the correct shards be done
automatically, without going to the entire cluster?
If not, is there a way for me to write a plugin to make this happen?
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.