Bug while promoting aurora read replica

Usually replication lag plays an important part on these kind of issues. Specifically we can have the following use cases:

• The replication lag between the primary DB cluster and the read replica has not reached 0 seconds yet. You can check the replication lag using the SHOW REPLICA STATUS command on the read replica. The replication lag needs to be 0 before promoting.
• There may be ongoing transactions on the primary DB cluster that have not completed replicating to the read replica yet. You need to stop any new write transactions to the primary before promoting the read replica.
• The data migration/synchronization between the primary and read replica cluster may still be in progress. You need to check the replication events and status to ensure migration is fully completed.
• There could be some issues detected during the pre-checks for promotion that need to be addressed first. You may need to check the RDS event logs or raise a support case with AWS to troubleshoot further.

Once the replication lag is 0 seconds and all data/transactions are fully replicated, the read replica cluster should be ready for promotion.