Can I ask how large your clusters are and how many records you’re reindexing? I haven’t been able to scale this to a cluster taking in several TB a day