An Azure managed PostgreSQL database service for app development and deployment.
Hi rishabh krishna,
Thankyou for reaching Microsoft Q&A!
If this suddenly started without any configuration change, the issue is usually network path, DNS resolution, or cached credentials rather than the database itself especially since other environments on the same subnets still work.
First, confirm that your client is really reaching the correct private endpoint. Because you are using a VPN + hosts file, verify that the hostname resolves to the current private IP. Azure Flexible Server private endpoints can change during maintenance or failover, and a stale hosts entry will cause authentication failures even though the server is reachable.
Next, test connectivity outside pgAdmin to rule out client caching:
psql "host=<fqdn> user=<user> dbname=postgres sslmode=require"
pgAdmin sometimes stores old passwords internally, so remove the saved connection profile and recreate it instead of just updating the password.
Also check these common causes that look like password failures:
Ensure you are connecting with the full username format username@servername (Flexible Server requires this depending on auth mode).
Verify that SSL is enabled. If SSL enforcement was updated by platform maintenance, non-SSL attempts can return authentication errors.
Confirm DNS from your machine while connected to VPN: nslookup <server>.postgres.database.azure.com The IP should match the private endpoint, not a public address.
Finally, review server logs under Monitoring > Server logs. If you see repeated auth failures but no network rejects, it confirms the client is reaching the server with wrong credentials format rather than being blocked by VNet rules.
Since password reset didn’t help and other environments work, I would strongly suspect hosts file/DNS mismatch or pgAdmin cached credentials those are the most common causes in private VNet + VPN setups.