Select Language:
If you’re changing your Multi-AZ RDS Postgres database from a db.t3.medium to a db.t4g.medium instance, and the process seems to be stuck after about an hour, here’s what you need to know.
First, understand that when you modify a Multi-AZ database, a series of steps happen in sequence. Initially, the standby replica is upgraded to the new instance class. After that, a failover occurs, promoting the upgraded standby to become the new primary. Finally, the original primary (which is now the standby after failover) is upgraded to the new class. This process explains why you’re still able to connect to your database and see data — the system is working through these steps.
As for how long this should take, it can vary. It might be slow, but it usually doesn’t get stuck. From your description, if it’s been 1.5 hours and the status remains “Modifying,” it’s a good idea to reach out to AWS Support for help. They can check if the upgrade process is truly stuck or just taking longer than usual.
You also asked whether the change might have failed and rolled back to the original instance type, db.t3.medium. Typically, if the modification fails, the status will display “Failed” or “Rolled Back.” If you’re still able to connect and perform database operations, it’s likely the process is still ongoing, and the instance class is being updated in the background.
In most cases, the system won’t let the primary and standby be on different classes in a Multi-AZ setup. So, if the status is still “Modifying” and your database is responding normally, give it more time. The upgrade should complete without issues. If it continues to be stuck, contacting AWS Support will help identify and resolve any underlying problems.
ChatGPT
Perplexity
Gemini AI
Grok AI
