So, I have a triangular router setup here, 3 sites, all of them connected to each other via GRE IPsec encrypted tunnels. That works like a charm.
I have set up OSPF with all three sites included, each site has OSPF p2p channels open to the other two sites. That works as well,
But now we come to the problem: ALL remote distances are at 110, regardless what the real path is. regardless what the cost are to get there.
In particular I have 2 sites inside Austria and one site in Amsterdam. The network packets still are taking the route from S1 in Austria via Amsterdam and back to S2 in Austria even though there is a clear route from S1 to S2, but unfortunately OSPF declared them to be the same cost.
Is there anything I can do to change the route costs and to enable the Total cost? Even if all paths are the same (which they clearly are not, since the interconnect from S1 to S2 is 15ms, whereas the path via Amsterdam is 60ms) OSPF should prefer the direct connection.
Hi,
Routing > OSPF > Interface Templates > Cost
Change the cost of the interface that should have lower priority
I did precisely that before I even started the post. The longer path has cost 20, the shorter path cost 1, still takes the longer path.
How do you announce the routes? In type1 or type2?
In a triangle you should not need to change any costs. leave everything as 1
short path (direct between sites) will have total cost = 1
long path (via other site) will have total cost = 2
Are you sure all your neighbors are in Full State, so that correct costs can be calculated?
Are you sure you have deleted any old static routes you maybe had from previous configuration without OSPF?
Hard to tell without export of OSPF and print of routes.
Ok, my apologies: I just saw that the host I was on was also in addition to being behind R1, it was also logged into zerotier and apparently some routing decisions are different then. I just confirmed without zerotier active the route is as expected.
Case closed, sorry for the excitement.
Ok, my apologies: I just saw that the host I was on was also in addition to being behind R1, it was also logged into zerotier and apparently some routing decisions are different then. I just confirmed without zerotier active the route is as expected.
Case closed, sorry for the excitement.