Update routing table in more scenarios

[whiskerz007]

Nothing was happening when a peer was disabled. This PR will make sure the peer is removed from WireGuard and the routing table is updated when a peer is disabled.

[whiskerz007]

@FossoresLP At this point, the routing table will match the state of the configuration. Currently, when the WireGuard interface is disabled, running wg show <interface> will still show the peers, still keep the running bandwidth usage, and keep track of the last handshake. Deleting the peers from the WireGuard interface would ensure wg show <interface> only shows the interface configuration and all peers will lose the running totals.

Should the peers be deleted from WireGuard when the interface is disabled in the configuration?

[FossoresLP]

I'm not quite sure what would be the better approach.
I don't think freeing up the resources would make much of a difference unless you are running many WG interfaces in disabled state, which I would say is a little bit odd to do.
What about the addresses - are they available for re-use when the interface is disabled?

[whiskerz007]

With the changes in this PR, when the WireGuard interface is disabled, the routes are removed. The IP address that is assigned to the interface is still configured. The script that updates the routes checks the list of allowed-ips from the output of wg show <interface>.

With the template rewrite, I tried to keep the same logic the was created before. If we need to redefine the logic, we should probably workshop some workflows to ensure the logic makes sense and the code functions as desired.

[FossoresLP]

I asked just out of curiosity, since this was requested before.

This PR looks good the way it is for now.

I'll merge it before the next release, but I don't think it's necessary to release it standalone since this is not a serious bug.

[whiskerz007]

With the current release, if you disable the WireGuard interface, the routes are removed (by the kernel). When you reenable the interface the routes will not update. I believe this is a bug that may not have been discovered.

[FossoresLP]

Ok, then I suppose a new release would be warranted. I'll get this PR merged and released in the next couple of days then.