on target oracle system the load was very high and therefore replication getting more and more delayed.
on source there normal day processing and there we was reaching cached changes 11Mio within 1,5Mio transactions and delay at this time of 6h
target system load was reduced but replication was not able to get fast enough back to sync.
After process on source system get very low during night, it was able to get back in sync. But this need about additional 8h.
We saw that on target for execution there was full table access necessary for "attrep_changes...." tables. An this was for tables with has higher processing load and took every loop longer
So my question is, if there is a configuration possible to have indexes for joining tables to optimizing processing?
