v1beta1 + v0.10.0

[EmilienM]

A few of us are actively looking at releasing CAPO v1beta1 (with a v0.10.0 branch) in the following weeks.
We would like to ask our community what they want to be in there before we cut the release.

I've dressed a list of things that I think would be required, I plan to maintain that list while the discussion is ongoing:

• ⚠️ Convert ServerMetadata from a map to a list #1828 (@EmilienM)
• Very likely OpenStackCluster.Status.Bastion.ReferencedResources is not filled up #1835 (@EmilienM)
• ✨ Move subnet options to SubnetSpec #1856 (@dulek)
• ✨ AllNodes security groups API #1826 (@EmilienM)
• Setting allowAllInClusterTraffic without managedSecurityGroups has no effect #1878 (@EmilienM)
• 🐛 Remove IdentityRef.Kind #1903 (@EmilienM)
• We need to remove omitempty from all values which aren't struct pointers #1850 (@mdbooth)

Please provide feedback and be involved!

[MaysaMacedo]

@EmilienM For the dual-stack support on NodeCIDR, we would need to create one additional subnet, IPv6 one this time and add it to the router. So this wouldn't be a large change if the Subnet uses a fixed address mode like dhcpv6-stateful. If we would like to let the user choose the address mode, that would be an additional API and could expand scope of this work.

On another topic, have we considered the following work for beta1? #1836

[EmilienM]

@EmilienM For the dual-stack support on NodeCIDR, we would need to create one additional subnet, IPv6 one this time and add it to the router. So this wouldn't be a large change if the Subnet uses a fixed address mode like dhcpv6-stateful. If we would like to let the user choose the address mode, that would be an additional API and could expand scope of this work.

Let's not do that, they can always do BYON, I don't think we can / should support every use case.

On another topic, have we considered the following work for beta1? #1836

Added. Thanks!

[mdbooth]

I don't think #1836 is a v1beta1 issue: I don't think fixing it can break any existing working configuration. Doesn't mean we shouldn't fix it, I just don't think it's a v1beta1 blocker.

[huxcrux]

Would be nice to have #1762 in v1beta1

I will check if #1687 still is a problem (hope to have it done today or tomorrow) to check if it's still a problem and in case if we can fixa a PR with a fix

[EmilienM]

Thanks for your input, I'll add it to the list but also wait for @mdbooth 's feedback on this one.