I am building an app similar to Uber for tracking vehicles. Since the update frequency is so high (accounting for several users), I want to know the general practices involved for making writes faster to mongodb collection.
I am maintaining a database to store historical location information from all vehicles but it is bound to grow very fast once we go into production. I need to get the list of vehicles closest to a point. For this should I implement a separate table (with one row per vehicle) which gets updated after every update or there is better/faster way to do this using the existing table?
Also I would like to know if there is a more efficient & faster database out there which can server my purpose.
Right now both the tables mentioned above have same structure
id vehicle_id loc timestamp
Links to articles with best practices are welcome
Is there a limit to number of inserts / updates mongodb can handle per second? In this video Curtis from Uber mentions that they discarded mongodb as they reached its limits soon after they went into production. I couldn't find any information on what they replaced it with.
I am expecting around 10000+ updates every second. Will this design still work for 100000 updates?