| === cpaelzer__ is now known as cpaelzer | |
| === tds7 is now known as tds | |
| === tds8 is now known as tds | |
| [15:52] <Odd_Bloke> blackboxsw: So I'm going to look through in more detail in a few minutes, but I think the key unresolved issue on #114 is the naming of the config option. | |
| [15:52] <Odd_Bloke> blackboxsw: I can't decide whether consistency with the other DSes or more accurate naming for this DS is more desirable. | |
| [15:53] <Odd_Bloke> (Though this is something of a false dichotomy: we could always rename those other config options too, while retaining support for the old spellings.) | |
| [16:32] <paride> blackboxsw, this is one: https://github.com/canonical/cloud-init/pull/254/files | |
| [16:32] <blackboxsw> thanks paride | |
| [16:33] <blackboxsw> +1 Odd_Bloke | |
| [16:34] <powersj> paride, what failed that required that change? | |
| [16:35] <paride> powersj, the nocloud-kvm tests: https://paste.ubuntu.com/p/XQ6Q8BHxXD/ | |
| [16:38] <powersj> paride, shouldn't this have been failing for years then? | |
| [16:38] <powersj> why now? | |
| [16:40] <paride> powersj, I was expecting this question :) And I don't have a definitive answer. Let me dig a bit more. | |
| [16:51] <paride> powersj, because the failing assert is new: https://github.com/canonical/cloud-init/commit/71af48df3514ca831c90b77dc71ba0a121dec401#diff-27f8cf430e53c95119b64a768e67e6e4R323 | |
| [16:51] <paride> added a comment to the PR | |
| [16:53] <blackboxsw> +1. yeah paride it was my bad :/ | |
| [16:54] <blackboxsw> and merged :) | |
| [16:54] <blackboxsw> yes because our cloud tests were contructing the version from yaml it interprets 19.10 as 19.1 float, so text comparisons in that case would turn out invalid. | |
| [16:55] <paride> blackboxsw, but it's yaml itself that has the concept of floats and strings | |
| [16:56] <paride> so I don't think it was a fault in your commit at all | |
| [16:56] <blackboxsw> paride: right, just at fault was that I only tested on *.04 series before landing :) | |
| [16:56] <blackboxsw> instead of *.10 :) | |
| [18:16] <blackboxsw> Odd_Bloke: how does the apply_secondary_network_config sound for ec2 multi-nic-secondary-ip PR #114? really, I'm up for any suggestion you feel is more tractable. | |
| [18:17] <blackboxsw> or understandable | |
| [18:43] <Odd_Bloke> blackboxsw: I liked the wording you changed to internally, perhaps `apply_full_imds_network_config`? | |
| [18:53] <Odd_Bloke> blackboxsw: FYI, I've opened up a few small PRs which should only take a couple of minutes to review: https://github.com/canonical/cloud-init/pull/255 https://github.com/canonical/cloud-init/pull/257 https://github.com/canonical/cloud-init/pull/258 | |
| [18:53] <Odd_Bloke> In case you're looking for another distraction. ;) | |
| [19:08] <blackboxsw> Odd_Bloke: thanks, renamed ds cfg for ec2 #114 and pushed. awaiting CI | |
| [19:09] <blackboxsw> will grab your reviews now | |
| [19:14] <garga> Hi! I have a question. I'm using cloud-init to deploy a CentOS 7 based VM at Azure and I need cloud-init to only configure eth0, it must ignore all other NICs present on the machine. Is it possible? I couldn't find anything relevant in the docs | |
| [19:29] <powersj> garga, are you using a custom image or something Azure provides? | |
| [19:29] <powersj> AnhVoMSFT, ^ | |
| [19:31] <Odd_Bloke> powersj: blackboxsw: I'm now +1 on https://github.com/canonical/cloud-init/pull/114. Are you +1 on dismissing rharper's review so we can land it? | |
| [19:31] <powersj> Odd_Bloke, done | |
| [19:32] <Odd_Bloke> powersj: Thanks! | |
| [19:32] <garga> powersj: custom image | |
| [19:40] <blackboxsw> Thanks for the review/land and bug update Odd_Bloke. So should we raise that bug now in ubuntu-release https://bugs.launchpad.net/ubuntu/+source/cloud-init/+bug/1866930 | |
| [19:40] <ubot5> Ubuntu bug 1866930 in cloud-init (Ubuntu) "[FFe] ec2 add support for configuring secondary NICs and secondary ipv4 and ipv6 addresses" [High,Fix committed] | |
| [19:40] <blackboxsw> since release team hasn't reviewed it yet? | |
| [19:41] <powersj> blackboxsw, yes please ping vorlon to ack it | |
| [19:41] <powersj> then upload | |
| [19:41] <powersj> and profit | |
| [19:41] <Odd_Bloke> We'll need to cherry-pick rather than new-upstream-snapshot, I think. | |
| [19:45] <powersj> I'm of the opinion that if we are going to SRU a whole version back right after focal releases, then why not do that now | |
| [19:48] <Odd_Bloke> powersj: Well, we don't have an FFE for most of the changes that have landed. | |
| [19:51] <Odd_Bloke> So I think we need to have a conversation with the release team about it, at the very least. | |
| [20:08] <blackboxsw> Odd_Bloke: put up the cherry pick branch https://github.com/canonical/cloud-init/pull/260 | |
| [20:08] <blackboxsw> for review | |
| [20:10] <blackboxsw> powersj: if we SRU everything, I think vorlon will likely say go through the whole SRU validation process which I don't know we have time for during this freeze period. I'd like to get funcationality public sooner in case there were a problem. Then we still have runway to fix an unexpected corner case before Focal is released. | |
| [20:10] <powersj> blackboxsw, that's fine | |
| [20:11] <blackboxsw> I think it's simpler/faster for the FFe, but we do have a minor wrinkle in our release process in that I'm also including dan's package build-deps change in #260 though I'm not sure if that'll cause some concern | |
| [20:30] <Odd_Bloke> blackboxsw: Including that is fine, I think, I've reviewed the PR. | |
| [20:36] <blackboxsw> thanks Odd_Bloke ok I queue to upload for review in the event that the FFe is accepted | |