nats-io/nats-server · error

Got an error loading %d index: %v

Error message

Got an error loading %d index: %v

What it means

While streaming log entries to a catching-up follower, loadEntry(index) returned an error other than ErrStoreEOF (e.g. store closed or corrupt WAL). The leader stops sending the current batch early rather than sending a gap; expected EOFs are silenced, so any log here is a genuine store read failure.

Source

Thrown at server/raft.go:3666

	if !leader {
		n.debug("Canceling catchup for %q, not leader anymore", peer)
		return
	}
	n.debug("Running catchup for %q [%d:%d] to [%d:%d]", peer, ar.term, ar.index, pterm, last)

	const maxOutstanding = 2 * 1024 * 1024 // 2MB for now.
	next, total, om := uint64(0), 0, make(map[uint64]int)

	sendNext := func() bool {
		for total <= maxOutstanding {
			next++
			if next > last {
				return true
			}
			ae, err := n.loadEntry(next)
			if err != nil {
				if err != ErrStoreEOF {
					n.warn("Got an error loading %d index: %v", next, err)
				}
				return true
			}
			// Re-encode with the lterm if needed
			if ae.lterm != term {
				ae.lterm = term
				if ae.buf, err = ae.encode(ae.buf[:0]); err != nil {
					n.warn("Got an error re-encoding append entry: %v", err)
					return true
				}
			}
			// Update our tracking total.
			om[next] = len(ae.buf)
			total += len(ae.buf)
			n.sendRPC(subj, reply, ae.buf)
		}
		return false
	}

View on GitHub (pinned to 3a66a489d2)

Solutions

1. Check WAL/stream store health on the leader for read errors
2. If the store is corrupted, the leader should step down so a healthy node serves catchup
3. Restart the leader process if the underlying file handles are stale

When it happens

Trigger: Thrown at server/raft.go:3666 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/6caeb2ac0a0366da. Report an issue: GitHub.