I had earlier gone through it. The overall flow helped, but I was specifically looking for how to iterate over values of an array field, after the latest change in field data structures, not for script executors. :(
On Wednesday, May 29, 2013 4:53:57 AM UTC+5:30, Otis Gospodnetic wrote:
Thanks for the link. Looks like this is what I was looking for.
I see that the purpose of the interface LongValueInDocProc is currently served to some extent by the class LongFacetAggregatorBase. I was implementing StringValueInDocProc in my code, and this Aggregator class is where I was stuck. Looks like I got the hint.
On Wednesday, May 29, 2013 10:19:58 AM UTC+5:30, dobe wrote:
Well - this looks *very* interesting. What kind of scale have you tested this at, and is it particularly sensitive to eg. high cardinality fields?
Hi, we use the pre 0.6.0 version of the plugin for about a year now on an index with about 3 billion docs. we have a field with about 1 billion different values where we use the latest and distinct facets. We plan to stage the 0.6.0 version in the next two weeks on a staging environment.
The facet was also in use with about 24 billion docs, so there should be no scaling problem.