DMS homogenous postgres RDS→RDS slows after initially working

Hi folks, I have a ~200GB postgres database on an m6g.xlarge instance that I am trying to migrate to a new instance. The end goal of the migration is to reduce our overprovisioned volume size from current 5TB to <400GB to reduce RDS cost.

The DMS migration task begins great and very quickly transfers some of the largest tables, around 100GB+ of data, then seems to slow to a crawl. Looking at the monitoring tab for the new RDS instance, the slowdown seems to correlate with the "EBSByteBalance" chart going from 100 very quickly to 0 and then being pegged at 0.

I have left the task running for a full day and see FreeStorageSpace continuing to decline at about 30GB/day (it is a straight line), so some additional data is being written. The tables that have data in them and are queryable in the new instance also seem to have replication working. But it's unclear whether the task is going to eventually succeed.

Is there a way to configure a new RDS instance to allow for a larger EBSByteBalance so I can complete the migration more quickly (if that is in fact the cause of the slowdown)? Or should I just wait...? At the current rate (30gb/day) it would take a few days for the rest to finish...