|A connection was successfully established with the server, but then an error occurred during the login process. (provider: TCP Provider, error: 0 - An existing connection was forcibly closed by the remote host.) (.Net SqlClient Data Provider)|
Finally I got an official alert from the Azure Service Health about a Service issues!
Good day guys,
If you encounter issues connecting Azure SQL Databases then you are not along. About 9 hours ago I started to see questions in the forums regarding issues to connect the service. My test confirmed that there is issue in specific region and I was impact by the issue in two of my developing servers in West Europe.
For the last 9 hours multiple reports arrives in the forums, but unfortunately Microsoft Azure subscription Service Health only alerted me about during the last hour or so!
Notice that the Azure status website report that there is no issue!
I will update you on any new information..
Update status 14:00 Israel time: Officially the issue still exists, but I do not see any issue in my subscription in the last hour. In face, from the time I got the alert about the issue, it seems like the issue was solved for me and connections are a bit slow but working.
Current status: The issue was fixed and closed. For more information scroll down to the "Summary of impact"
In the meantime, My conclusions are that we are better follow the forums and online reports in the communities if we want to get the real time alert about potential issues. At the same time, if you do encounter an issue, then please report it in your community communication channels like the Microsoft Forums and you can even join us at the Microsoft Wiki group on Facebook and report there. There are over 70 Microsoft MVP, multiple Microsoft employees, and tens of active members which manage azure accounts. Your report can help make the difference!
Summary of impact
Between 23:51 UTC on 05 June 2020 and 09:55 UTC on 06 Jun 2020, multiple customers encountered issue to connect SQL Database / SQL Data Warehouse in West Europe region. New connections to existing databases in this region may have also resulted in an error or timeout. Retries may have succeeded.
Microsoft team determined that some instances of a gateway service became unhealthy and were unable to process connections. Since the connections are done using load balanced across the instances, customers would have experienced issues if their connection was directed to the unhealthy instances, while in other cases the connection was redirect to healthy instances - this explain why some attempt to connect succeed while other failed.
Microsoft confirmed that after extensive investigation, they mitigated the issue by restarting the unhealthy instances of the gateway service, and validated that traffic was flowing as expected, mitigating the issue.
Note: The information in the summery came from Microsoft official report in the azure Service Health | Health history
Note: (last update Jun 7, 10:50 IDT) during the entire issue I did not noticed any alert in the Azure status website, which as much as I know reported that there is no issue during the entire event! Moreover, until this point in time the Azure status history site does not include report on the event. I assume it will be updated after the weekend.
With that, I think that I can close this followup. At least for my clients it is not needed anymore. Have a great day guys and stay safe