At present, the cluster cannot be recovered. I cleaned up the data and redeployed the cluster once. When I found the amount of empty data, the time taken to view the query and mutation process from the retal interface has changed from more than 4ms to more than 20ms. Any suggestions to fix。
