> This surprisingly brought the entire cluster online.
@Peter, that's both good and interesting news. If this is a viable workaround, we should probably just document it as part of the upgrade. I'm still on the fence on how to actually solve the bug in a robust, elegant, way.
> This surprisingly brought the entire cluster online.
@Peter, that's both good and interesting news. If this is a viable workaround, we should probably just document it as part of the upgrade. I'm still on the fence on how to actually solve the bug in a robust, elegant, way.