Microsoft
Microsoft
MicrosoftThanks twhitecotton for your feedback. I would love to learn more about the DNS cleanup issue which you are referring to, which is not normal and not a widespread known issue which we are aware of. We try to document all the known issues and bugs along with feature releases below for easy reference and visibility for our customers.
https://docs.microsoft.com/azure/mysql/flexible-server/whats-new
Regarding your server shown as Healthy and Available in portal and no automatic failover, let me expand on how failover detection works in Flexible Server. As long as MySQL server is up and running and ready to accept connections, you will see the status in Azure portal as healthy and there is no reason for service to automatically failover. In your scenario, it seems the issue was on the networking path between client and server due to DNS records and in this scenario, the server will not failover and it is by design since it may still not help you to connect to standby as well if the issue is on networking or DNS path. I can understand from your perspective, failure in connectivity is similar to unavailability as end outcome results in application downtime but technically, these are different from system design perspective as the server is available and ready to accept connections but the server cannot be reached due to errors on networking or DNS path. For transparency and clarity, we have documented the same below
Having said that, I can imagine you may not be happy with my response as you would and should expect better. Thanks for holding the bar high for us. I would like to learn more about the issues you are experiencing and work on improving your service experience end to end. I would be thankful if you can email us Ask Azure DB for MySQL <AskAzureDBforMySQL@service.microsoft.com> with your support ticket numbers for us to investigate and learn more.
- Parik
