I would like to completely control my network transportation by defining the packages path. According to this I can't do it unless I have control over all the PCs/ routers in its way. Suppose, I can buy/ own different PCs/ routers in specific places (GEO locations). Is it sufficient to fully control and determine the packages path?
1
-
That is exactly the opposite of what IP is for. What you are asking about is something like the old circuit-switched networks. Packet switched network, like IP, are designed so that you simply do not care the path packets take. The original premise was that the path can dynamically change based on conditions (congestion, natural disaster, terrorism, nuclear war, etc.). In any case, if you mean the public Internet, then, no, you do not have any control. You could make your own large network and lease the lines between sites, but that is very, very expensive.– Ron Maupin ♦Commented Oct 1, 2019 at 20:12
Add a comment
|
1 Answer
Suppose, I can buy/ own different PCs/ routers in specific places (GEO locations). Is it sufficient to fully control and determine the packages path?
Unless all these routers are directly connected to each other, no. Each hop decides on the egress route on its own volition, its routing table, policies or whatever. You can't influence that on routers that you don't control.
As a compromise, you could use tunneling to create an overlay network on top of nodes you control. You can then control routing through the tunnels (inside routing) as you desire. You cannot control the routing of the encapsulating packets themselves (outside routing), however.
Source routing is pretty much generally deprecated and completely ignored on the Internet.
Additionally, trying to artificially control packet routing is moot at best, normally. If you provide each router with the best routing table possible (which is pretty much the norm) there's nothing to improve there.
