Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good news: CI passed
Bad news: CI takes 20+ minutes longer with 0.2

We should look at parallelizing the conformance tests, there's no reason they need to be sequential (other than the logs are easier to follow)

I see two common errors, both benign:

1. The "unable to sync" errors with the caused "stop processing SubReconcilers, without returning an error" needs to be suppressed. It's an expected case and not an actual error. The cause is the case where the webhook was intercepting the request that was applying the binding. I think this is what you linked to, deep links into large logs get wonky.
2. optimistic lock failures when updating a resource that was mutated under us.

Was there another error you were seeing?

nope, I wasn't sure if those were intentional or not, sorry for the confusion.