Interesting stuff, definitely the kind of real world optimization that happens when you’re able to look at actual access characteristics rather highly abstracte

Interesting stuff, definitely the kind of real world optimization that happens when you’re able to look at actual access characteristics rather highly abstracted models. At one level you could just be saving the results into a hashtable to bubble them out again, but at the extreme end the lookup is actually spread across multiple machines in a cluster, so it’s more like throwing the query to the proper one with the right chunk of the index, along with instructions about where the final RPC is supposed to go with the result.