Add multiples routes options

[agustim]

Hi @alecbcs @dukzcry
This PR is based on #62 but you can set multiple routes. I've move route field to routes section where you define CIDR and node to route, like this:

interface:
  ...
peers:
  10.0.0.1:
    id: Qm1....
  10.0.0.3:
    id: Qm2...
  ...
routes:
  192.168.3.0/24:
    ip: 10.0.0.3
  0.0.0.0/0:
    ip: 10.0.0.1

[yinheli]

Sounds great, ping @alecbcs

[bmullan]

I have a question about this multiple routes capability.

The first comment's example says:

interface:
...
peers:
10.0.0.1:
id: Qm1....
10.0.0.3:
id: Qm2...
...
routes:
192.168.3.0/24:
ip: 10.0.0.3
0.0.0.0/0:
ip: 10.0.0.1

I have 2 Ubuntu 22.04 VMs (called vm1 & vm2)

In both VMs I've installed LXD.
in both VMs I've created an LXD container named (cn1).

When Logged into ether cn1 container I can ping anything on the Internet and I can ping "that" LXD Container's VM host.

On each VM I installed hyprspace.

On vm1 my yaml config file looks like:

interface:
name: vmhs
id: QmTB1pXv51fJXLyQF8JiwPQNFZCRj8vSVQAJrGPDBdpG5k
listen_port: 8001
address: 172.16.1.1/12
private_key: !!binary |
... key value

peers:
172.16.1.1:
id: QmTB1pXv51fJXLyQF8JiwPQNFZCRj8vSVQAJrGPDBdpG5k
172.16.1.2:
id: Qmaea2NerBfsCunQVnck34bF9mNLrrDYZdLV1LydcL9PKz
routes:
10.169.80.0/24:
ip: 172.16.1.1
0.0.0.0/0:
ip: 172.16.1.1

where:
On VM1 all LXD containers connect to their VM Host via an lxdbr0 bridge whose address is 10.169.80.1.
So I was assuming.. the above ROUTES section meant routing any traffic from the 10.169.80.0/24 network (ie the VM1 local LXD container's network) to the VM1 local Hyprspace VPN tunnel interface 172.16.1.1**

On VM2 I have the same setup and that Hyprspace config as on VM1 but the ROUTES config again is configured to Route VM2's local LXD container network (10.2.13.0/24) to VM2's own Hyprspace VPN Tunnel IP (172.16.1.2) where I was assuming it would get forwarded to VM1's VPN Tunnel interface (172.16.1.1) and on to the VM1 LXD network & the target LXD Container.

From VM1 I can PING VM2's Hyprspace VPN Tunnel interface (172.16.1.2)
I can not PING from VM1 to VM2's LXD bridge interface IP !

So hyprspace is not routing traffic from VM1's LXD container network to any of VM2's LXD containers.

Whether its LXD or Docker containers they run on their "Host" on a private (often a 10.x.x.x) network.

Isn't this use-case scenario what the hyperspace "multiple routes capability" is for or am I misunderstanding it's function?
For instance, I can not ping from VM1's CN1 container to VM2's CN1 container?

thanks
brian

[bmullan]

@alecbcs @dukzcry

My First Question is:

Did the Hyprspace Routing capability ever get implemented in a Release version?

If Yes... then my Second Question concerns interpretation of the "routes" section of the config:

If 2 Servers on 2 separate Cloud networks use Hyprspace to create a VPN tunnel to connect to each other, I thought the PEER list would look something like the following on one of the 2 Hyprspace Peer Nodes.
_

Peers:
172.16.1.1:
id: QmTB1pXv51fJXLyQF8JiwPQNFZCRj8vSVQAJrGPDBdpG5k
172.16.1.2:
id: Qmaea2NerBfsCunQVnck34bF9mNLrrDYZdLV1LydcL9PKz
routes:
10.169.80.0/24:
ip: 172.16.1.1
0.0.0.0/0:
ip: 172.16.1.1

_

If 10.169.80.0/24 is the IP for a local Docker -or- LXD Container Network on one of the Server Nodes with the above Peer Config... then is the above telling the Server's Hyprspace to:

• Route traffic FROM the local Node's Container Network to/through the Node's_local_ Hyprspace VPN Tunnel End Point (TEP) IP 172.16.1.1

-or- does it mean...

• Route Traffic TO the 10.169.80.0/24 Container Network on the "other" Peer whose Hyprspace local TEP IP is 172.16.1.1 ??

Thanks
Brian