Saw this while debugging another issue on a support call for a 3.6.0 cluster, but haven't been able to reproduce myself on the sdk. It seemed like the metrics in the Hydrator pipeline were from the previous run instead of the current. For example, we ran the pipeline once and it wrote 2 records, but the metrics on the UI showed 0. Ran it again on 6 records and the metrics showed 2.
Not sure if this is a UI or a metrics issue.