Today the HealthPeriodicLogger only emits logs on the currently-selected health node, but if the cluster is sufficiently unhealthy (i.e. preflight checks failing) then there may not be a currently-selected health node. Or it may be that a node cannot join the cluster so the cluster thinks it's healthy, and the excluded node is not the currently-selected health node so it cannot report otherwise. In these cases the preflight health checks are failing, and in that case nodes should be able to make the unilateral decision to report that they are unhealthy.
Today the
HealthPeriodicLogger
only emits logs on the currently-selected health node, but if the cluster is sufficiently unhealthy (i.e. preflight checks failing) then there may not be a currently-selected health node. Or it may be that a node cannot join the cluster so the cluster thinks it's healthy, and the excluded node is not the currently-selected health node so it cannot report otherwise. In these cases the preflight health checks are failing, and in that case nodes should be able to make the unilateral decision to report that they are unhealthy.