We have the pega api response issue, the response is 572ms before, after that we downgrade the cpu/memory to half of that and later we change to related api nodes from 6 to 3, total nodes from 15 to 7, then we found the response time is go to 3s+, so we stop all the change back to full cup/memory and nodes to original total 15/ api 6， but the response can’t go back to 500ms+, we want to clear if such rollback we need clean up some confirmation with db/table side, or any steps we did’t do will make such a result?  We confirm testing the api with the same env. the same data volume. Many thank.