Summary issue:
Config cluster with 5 SCs.
Starting cluster: SC5 is active, SC4 standby and [SC1-SC3] is quiesced. SC1 and SC2 are promoting those nodes to become ACTIVE (by rde component). SC5 is answering the take-over request from SC2 first and reject this request. Then SC5 is rebooted (manually) before answering the request from SC.
After SC5 is rebooted. SC4 is also promoted it self node to become the ACTVE. SC4 send the take-over request.
Unfortunately the node become ACTIVE is quiesced SC1, not standby SC4. So all synced data from ACTIVE may be lost
Diff: