Harbor high availability

[lixiaopengy]

I checked the official high availability example and found that two harbors share a database, and the back-end image storage adopts shared storage. In this case, the replication task between the two nodes will not be available. Because I checked the source code, it shows that the trigger condition of replication task is that there is no blob registration data in the instance node before synchronization. Do you have more relevant high availability documents to learn from