I’ve been trying to research best practices in this area and am finding some conflicting recommendations. In our use case we have multiple indexes and need to combine the use of single and multi readers to target content as the user selects one search area or areas over others. We’ve tried SearcherManagers and warming strategies in an effort to reuse Searchers and avoid rebuilding readers over potentially large numbers of indexes, some of which can be fairly large. But these have had mixed results and have introduced some errors. Can someone comment on current best practice and what might work well for our use case? Thanks, Scott
