cloud-init-dev team mailing list archive

[Scott Moser <smoser@xxxxxxxxxx>]

Mike,
Yeah, thats true.
The fix here does verify that the content is better than it was in that bug report though.

We need full round trip rendering and reading to do the whole thing the right way.

That said, the network config that was provided in the bug, and I think correctly simplified here, has static routes that are not attached to an interface. There is no specific link between the 
'route' type and the subnet that it should be off of.

we could arguably work backwards and find the right subnet to hang the static route off of, but at the moment we dont do that.   Additionally, there could well be two devices with an 'address' that would contain the route's 'gateway'.

Again, I agree that this test is minimalistic.
You have thoughts on what the route info should look like?

(we should probably move this discussion at least to the bug).

-- 
https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/313475
Your team cloud init development team is requested to review the proposed merge of ~smoser/cloud-init:bug/1649652-curtin-routes-rendered into cloud-init:master.