What is the best way to store GPS tracks in a database?

Question

satellite navigation

A theoretical question to consider.
There are many GPS tracks, each of them has many points. What does a lot mean? Since the question is theoretical, I will not name specific figures, but potentially there can be up to several million here and there. That is, multiplying one by the other, we get a trillion.
What is the best way to organize storage in a database? One friend told me, they say, we are making a table of tracks and a table of points. All points fall into it en masse, and when you need to get a specific track, points are selected by track id and sorted in order. It seems to me that this is not very effective and will work great on small volumes, but with their growth it will fall very quickly.
Store the points of each track as a blob? Even more strange.
Looking towards NoSQL?
Maybe there are standard schemes?

Answer 1

PostgreSQL + PostGis. Store as regular coordinates of the desired precision. A selection of any complexity is possible.
Now my coordinate base occupies more than 70GB, the selection of coordinates inside the desired polygon takes about 6ms on a regular hdd.
If desired, you can cram the tracks in the form of a multiline from which it is then easy to pull out individual points
PS the PostgreSQL + PostGis bundle uses OSM, where the unpacked planet base easily exceeds 500GB

Answer 2

You can use geodata in relational databases. For example in MS SQL