If i use Prometheus remote write integration in a high-availability (HA) configuration, how can i make sure my Prometheus servers not send multiple copies of the same metrics

If i use Prometheus remote write integration in a high-availability (HA) configuration, how can i make sure my Prometheus servers not send multiple copies of the same metrics?

A recent suggestion was a tool such as GitHub - beorn-/highlander: Reverse proxy handling HA for prometheus server groups remote_write feature

Also some remote write endpoints (e.g. Thanos) already handle deduplication for you.