Recently used nested R-Trees for in memory geospacial querying for a game (millions of atomic updates a second which would be a pain to manage sharding as items move around the world).
For example as a tree gets too big I split it at the hot spot and run the job processing for each tree on separate threads.
Few years ago implimented a simple tree for a survey product I built. The pathing in the survey is stored in the tree and I use the tree to find loops (which are invalid) in the survey.
Recently used nested R-Trees for in memory geospacial querying for a game (millions of atomic updates a second which would be a pain to manage sharding as items move around the world).
For example as a tree gets too big I split it at the hot spot and run the job processing for each tree on separate threads.
Few years ago implimented a simple tree for a survey product I built. The pathing in the survey is stored in the tree and I use the tree to find loops (which are invalid) in the survey.