Hi all, In our production app, we iterate around

5k users

where for each user it runs typesense

multi_search

with atleast

4 search_requests

then we log each request where the

search_time_ms

exceeds

2 seconds

, from the logs, some requests from 1 collection are around

6 seconds

then I tried to replicate these on my local using the same search parameters but somehow they all return

100ms

search_time_ms

, anyone encountered that same case? thank you

Check if there is concurrent indexing happening during the time those search queries were run.