Microsoft
MicrosoftHi - thanks for the write up this surely solves one challenges out of the 2 for DR with Azure postgres .
The solution recommened should help the services point to Primary or secondary without being have to repoint or reconfigure the application , but once an replica is promoted as primary in the secondary region , once DR event is resolved or drill is completed goal would be switch everything back to the primary . The DNS change can be made but now that replication is cut and secondary (which is the current primary) has updated data , the only option left for us is to delete the old primary , re-create it and again sync it with the current primary (in DR region) and repoint . and once thats done , again to make our application DR compliant , we have to delete the server in the secondary region and initiate a fresh DR sync with the primary . Thats the problem we are trying to solve , and this is solved in Azure SQL with Failover groups and Azure dosent seem to have any such features for any flavors of postgres . We didnt have it with Single server and hoped to see it in Flexi , it didnt have it and neither does the new Cosmos Postgres (Citus ) have it .
