> I like Jason's suggestion, although I would want to check for "y" as well.
>
Will fix. I was following the use of T in a few other places, so switching the others should be done in the branch that is adding in better opt parsing
> I'm a bit concerned about the way which _on_set_network_host is being called. It seems to imply that all hosts are in a single HA network. Is it possible that there are multiple, disparate, HA networks?
My current mental model says HA networks belong to all hosts. I think this is the most common use case. That said, I can change the init host to look for ha networks that already have an ip on this host, and only on_set_network_host there, then I can also call it from the associate call in _get_dhcp_ip. That should make the network only show up on the host if it is being used.
On Jul 7, 2011, at 3:13 PM, Mandell wrote:
> I like Jason's suggestion, although I would want to check for "y" as well.
>
Will fix. I was following the use of T in a few other places, so switching the others should be done in the branch that is adding in better opt parsing
> I'm a bit concerned about the way which _on_set_ network_ host is being called. It seems to imply that all hosts are in a single HA network. Is it possible that there are multiple, disparate, HA networks?
My current mental model says HA networks belong to all hosts. I think this is the most common use case. That said, I can change the init host to look for ha networks that already have an ip on this host, and only on_set_network_host there, then I can also call it from the associate call in _get_dhcp_ip. That should make the network only show up on the host if it is being used.
Vish