The origin issue is from SDB-3484. When a customer wants to update its pull-secret, we find that sometimes the insight operator does not execute the cluster transfer process with the message 'no available accepted cluster transfer'. The root cause is that the insight operator does the cluster transfer process per 24 hours, and the telemetry does the registration process per 24 hours, on the ams side, both the two call /cluster_registration do the same process, so it means the telemetry will complete the cluster transfer before the insight operator. 

4.12

Always

1. Create two OCP clusters.
2. Create a PSR that will help create two 'pending' CTs. The pending CTs will be accepted after ~6 hours.
3. Wait for ~24 hours, check the PSR, and check the logs in IO, and also check the pull-secrets in the clusters.

The PSR is completed, but there is no successfully transfer logs in IO, and the pull-secrets in the clusters are not updated. 

The transfer process is executed successfully, and the pull-secrets are updated on the clusters.