In the example, they're creating a mapping for "my-type" with a "message" field. Then when trying to match a document, they are calling this request: "/my-index/my-type/_percolate". However, if I replace "my-type" with "this-doesnt-exist", I get the same results back. When I replace it with "products" (which does exist), I get the same results back, but it takes longer.
How exactly is "my-type" used when searching for matching documents?
The type option is used the tell the percolator how the index the document that you're percolating into a temporary memory index (that is used the evaluate your percolator queries against). For that it uses the mapping (and its configured field types, index settings, analyzers), which indicate how the field values in the document to be percolated should be analyzed when indexing.
If a type doesn't exist default field types and analyzers will be used. Perhaps your products type has different field types or analyzers is configured that cause the percolator queries to require to do more work. Can you share the products mapping?
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.