I have two different indices with same data but different index mappings(one has around 6000 indexed fields, and the other has around 70). I need to compare the querying speed on these two indices. What is the best way to benchmark? As in I have following issues:
- How do I need to disable caching in queries. (I know real world would have caching enabled but I would just like to see the effects of an improved mapping)
- Does the number of replica for the indices matter ?
- what other parameters do I need to ensure are same before proceeding with the benchmarking?