Silvio Moioli

@richard-cox: @MSpencer87 discovered another side of this bug and I feel like needing your help to figure out why it happens. Assuming the same reproduction instructions above, `too old resource...

@MSpencer87 given the last comment from @richard-cox are you OK considering this done and addressing the remaining problem you found in the follow-up issue https://github.com/rancher/dashboard/issues/10540?

@MSpencer87 can you report test results here please?

@snasovich I need you or somebody from your team to take a look at this PR, can you point me to the most appropriate person? Thanks in advance

> how extensively has this been tested? @jakefhyde I have been testing this myself, QA instructions are ready. One customer tested it in a debug image but reported a regression:...

> @moio yoy may need to check to see if the cluster even needs a dialer, see: > > https://github.com/rancher/rancher/blob/5314106f746c0c62e95d54922a2dfd116d2f6f68/pkg/dialer/factory.go#L189-L192 > > there are specific cluster types where we don't...

> [...] Implementing your proposed fix [for managed clusters] introduces a chicken-and-egg situation for that case where we are deploy the `cattle-cluster-agent` which runs the `healthsyncer` which means the `Ready`...

@Oats87 can you take a look at the new version of this change which is radically different and (to my understanding) follows your comments? **Note this is best reviewed commit-by-commit.**...

@Oats87 related: what would be your opinion about extending this approach to `statsaggregator` (see 460eb9ac9) and potentially other controllers?

Thanks for the reviews! Minor details still up to discussion. I will merge as soon as either @erkki or some other user can confirm this helps out there in the...