Question

I started configuring Solr based off of the original example and so there were many random fields that they included for the example.

I might use one or two of them but for the most part they are all just sitting there unused (nothing in my data config mentions them or stores data to them.

Does this create any impact on performance ( even a small amount ) & should I just go ahead and delete them or does it not matter?

Thanks guys

Answer 1

I recommend deleting the fields and associated unused types. Possibly no performance impact, but there are some dynamic fields and copyFields which may cause confusion down the line. Or just build a new config directly instead (you only need two files to start: schema.xml and solrconfig.xml).

If you are doing it by trimming unused fields, make sure to keep or change whatever your uniqueKey points at, whatever your ‘df’ default field is and _ version_ (no spaces with underscores).

The last one is requires for real-time get and updateLog, which are enabled in example’s solrconfig.xml as well. Easier to keep _ version_ than to try removing all those things.

(Update Jan 2017: There is now a presentation video, specifically working through examples and how to clean them)