Nathan Sweet
Nathan Sweet
I agree with much of @huntergregory's assessment, and even more implementation hurdles could be discussed (such as distributed logging in large clusters not being scaleable, etc.), which I am happy...
> It is not an authoritative source of truth for the network policy spec, but a way to verify the netpol spec won't break the cluster. You do not get...
@npinaeva > That is the only requirement for this field, and the only reason to make it an API field: it will be required to be considered conformant. I don't...
@npinaeva > Right, I have not updated the description for a while, fixed now. I think the easiest solution for upstream is to say that the dry-run mode action is...
@jrajahalme > I did not try stacking the indices see how much it would help. I think the port-protocol indices should be used first. This makes the logic easier to...
Hit in: https://github.com/cilium/cilium/actions/runs/9666057377/job/26718389249 PR: https://github.com/cilium/cilium/pull/33242