luwak
luwak copied to clipboard
Three possible bugs in DocumentBatch
Hi, I started to work with LUWAK for percolator functionality in SOLR. I use ParallelMatcher with SimpleMatcher inside, and DocumentBatch with 3000 docs per batch.
During using LUWAK, I noticed 3 things (maybe bugs):
-
First issue with solution you can read here: https://github.com/flaxsearch/luwak/issues/171 It's relevant for big size batches.
-
Second issue with temp solution and optimal suggested solution you can read here: https://github.com/flaxsearch/luwak/issues/172 It's relevant for big size batches with different fields between docs
-
Also I noticed when I work with DocumentBatch, I must provide _luwak_id field. It's not a problem, but maybe to metion about it in README file?