Please closely monitor your metrics (especially latency and disk IOPS) after you change the instances or decommission replicas
Please update the status in the "Status" column -> if you already tried our recommendation and it doesn't work, please state it in the "Status" column too