Lately I see a lot of skipped scheduler searches in my search head cluster (3) with the reason "Your maximum number of concurrent searches has been reached. usage=3104 quota=3000 user=admin".
If I view from DMC console, I see not more than 200 searches run for each search head summing up to almost 600~700 searches for admin user in the last 4 hours.
I do not see any active running searches neither in the search head job manager nor on the Scheduler activity in the DMC.
However I see a lot of deferred jobs for one of the search head in DMC console.
Do you think rolling restart of search heads will help here as I do not see any search related issue here.