stale routing for host only network

[xenoterracide]

I have these 2 boxes (configured by vagrant), but I have a problem, when I start distributed and ping 172.28.128.2 often the network isn't working, but if I then go into oracle and ping 172.28.128.3 it'll find that and both networks will suddenly be able to talk to each other (once saw ip neigh in a DELAY state)

Code: Select all

config.vm.box = "chef/centos-6.5"

config.vm.define "oracle", primary: true do |oracle|
    config.vm.network "private_network", ip: "172.28.128.2"
    config.vm.network "forwarded_port", guest: 1521, host: 1521

config.vm.define "distributed", autostart: false do |distributed|
    config.vm.network "private_network", ip: "172.28.128.3"

Code: Select all

[root@localhost ~]# ip neigh
172.28.128.1 dev eth1 lladdr 0a:00:27:00:00:01 REACHABLE
10.0.2.2 dev eth0 lladdr 52:54:00:12:35:02 REACHABLE
172.28.128.2 dev eth1 lladdr 08:00:27:e3:d8:31 DELAY

this is on 4.3.12 on OS X Host, what could be going wrong?

[Perryg]

Unfortunately we don't support Vagrant.

[xenoterracide]

I don't think vagrant is realistically a direct part of the problem, it's still a host only network, the route obviously exists, but why does it get stale. If I started not using vagrant to bring them up and down, then they'd just be virtualbox images... with hostonly adapters having network problems.

[Perryg]

It might be but I have never seen it and given the configuration changes I see it would be difficult at best to diagnose.
Now if you can produce this without the use of vagrant it might be able to figure it all out.