nats-io/nats-server · error

Falling behind in health check, commit %d != applied %d

Error message

Falling behind in health check, commit %d != applied %d

What it means

isCurrent detected commit > applied and the gap did not shrink after polling for up to 10x1ms — the state machine is not consuming committed entries. The node is flagged not-current (hcbehind set), which affects health reporting and can trigger leader replacement.

Source

Thrown at server/raft.go:2196

	if startDelta := n.commit - n.applied; startDelta > 0 {
		for i := 0; i < 10; i++ { // 10ms, in 1ms increments
			n.Unlock()
			time.Sleep(time.Millisecond)
			n.Lock()
			if n.State() == Closed {
				n.debug("Node closed during health check, returning not current")
				return false
			}
			if n.commit-n.applied < startDelta {
				// The gap is getting smaller, so we're making forward progress.
				clearBehindState()
				return true
			}
		}
	}

	n.hcbehind = true
	n.warn("Falling behind in health check, commit %d != applied %d", n.commit, n.applied)
	return false
}

// Current returns if we are the leader for our group or an up to date follower.
func (n *raft) Current() bool {
	if n == nil {
		return false
	}
	n.Lock()
	defer n.Unlock()
	return n.isCurrent(false)
}

// Healthy returns if we are the leader for our group and nearly up-to-date.
func (n *raft) Healthy() bool {
	if n == nil {
		return false
	}

View on GitHub (pinned to 3a66a489d2)

Solutions

1. Profile what is blocking the apply loop (slow consumers, blocked callbacks, CPU starvation)
2. Scale or tune the upper layer draining the apply queue
3. Check for paused apply queues that were never resumed

When it happens

Trigger: Thrown at server/raft.go:2196 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of nats-io/nats-server@3a66a489d2 (2026-09-02). Data as JSON: /api/errors/930695c2bd5eb962. Report an issue: GitHub.