cloud-init team mailing list archive

[Frederick Lefebvre <fred.lefebvre@xxxxxxxxx>]

The thread at openstack got some interesting feedback initially and then
quickly converged on using a well known address. That is admittedly the
easiest solution to implement for clients but I feel there is a missed
opportunity there, to offer a long term solution to the disparity we
observe today when dealing with metadata endpoints on various cloud
platforms.  Requesting a reserved multicast group to discover metadata
endpoints might open the door to more uniformity.  The TCP connection would
still be established with an unicast address but, at least, the client
would not have to know, before hand, of the addresses for all platforms.

Frederick


On Wed, Mar 11, 2020 at 2:40 AM Slawek Kaplonski <skaplons@xxxxxxxxxx>
wrote:

> Hi,
>
> I’m OpenStack Neutron developer. Since some time we have proposed spec [1]
> about Metadata over IPv6 in IPv6 only networks.
> This spec proposes to use fe80::a9fe:a9fe IP address as it is equivalent
> of 169.254.169.254 in IPv6 link-local subnet.
>
> According to [2] I know already that cloud-init can be configured to use
> any metadata urls with “metadata_urls” so we can use such IPv6 address (or
> any other) and can be configured to be used in cloud-init on guest vms.
> But as cloud-init is one of the biggest users of Metadata service, I would
> like to ask cloud-init community what do You think about it and if this
> will work for You.
> Any feedback here or in the spec's review will be appreciated :)
>
> [1] https://review.opendev.org/#/c/315604/
> [2]
> https://github.com/canonical/cloud-init/blob/master/cloudinit/sources/DataSourceOpenStack.py#L59
>
> —
> Slawek Kaplonski
> Senior software engineer
> Red Hat
>
>
> --
> Mailing list: https://launchpad.net/~cloud-init
> Post to     : cloud-init@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~cloud-init
> More help   : https://help.launchpad.net/ListHelp
>