Hi SOLR Community, I've been looking at performance tuning solr's GC lately. I found this helpful article on the matter. https://cwiki.apache.org/confluence/display/SOLR/ShawnHeisey
One thing the article does not address is the GC1's ability to use string deduplication: https://blog.gceasy.io/2018/12/23/usestringdeduplication/#more-2861 Are there scenarios where this is advisable for SOLR? For example, if my cluster has many requests which come in with the same basic filters (things like site, username, type filter, etc...) would you expect the performance gain to be noticeable as SOLR may not be frequently recreating the same dictionary keys, etc? Or since there is likely some overall loss in performance, would you expect the loss to be likely to outweigh any of the gains? Thanks! Stephen
