works well for approximations. I am not sure our clients will like that. There could be many corner cases (e.g. leap year, DST, etc...). Also, does not account for the year. Its just 182d relative to today's date if I understand this right.
Thank you David. I read that thread you shared yesterday (see my comment in that ).
The pre-processing workaround is a smart idea, but unfortunately not feasible for us. What we attempted to do was to run a script that calculates whether it is 1st or last 6 months of the year. This takes forever if the index is large and leads to OOM error since the query performs a linear scan on the index along with some processing.
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.