I'd like to filter / facet on a text field that contains a code ( eg. abc,
ab3, abz ... ). Do I need to set this field to { "index" : "analyzed" } (
or more appropriately { "index" : "not_analyzed" } since i'll want to
filter on the exact string. )??? What about other field types ... numeric,
dates, etc.
Filters seems to work fine on my development dataset ( hundreds of
thousands of docs ) when they aren't indexed, but I've yet to test a large
dataaset ( hundreds of millions of docs ).
I feel silly asking this question, but my other database experiences aren't
mapping to ES as I don't have a good handle on the underlying technology
yet.
You are correct, if you want to filter on the exact value of a string
field you should set it to "not_analyzed". This is only needed on
string fields due to tokenization of the string during indexing when
not set.
I'd like to filter / facet on a text field that contains a code ( eg. abc,
ab3, abz ... ). Do I need to set this field to { "index" : "analyzed" } (
or more appropriately { "index" : "not_analyzed" } since i'll want to filter
on the exact string. )??? What about other field types ... numeric, dates,
etc.
Filters seems to work fine on my development dataset ( hundreds of thousands
of docs ) when they aren't indexed, but I've yet to test a large dataaset (
hundreds of millions of docs ).
I feel silly asking this question, but my other database experiences aren't
mapping to ES as I don't have a good handle on the underlying technology
yet.
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.