Today, our Azure Web App hosted in North Europe experienced an unexpected 30-minute downtime due to a re-imaging process. After reaching out to Microsoft support, they informed me that this was classified as 'unplanned maintenance' and suggested using 'Application Initialization' to prevent cold starts. Since most of our application requires users to log in, I'm concerned about how to mitigate similar risks in the future. Currently, when we deploy changes, we have to manually visit around 100 pages to warm up the site before switching slots. I'm looking for effective strategies to address this issue. Any thoughts would be appreciated!
2 Answers
Consider scaling out to two instances. This way, if Azure needs to do maintenance, your app stays up and running without any downtime.
I believe there's no guaranteed uptime for web apps unless you have at least two instances running. This is a common pitfall that can lead to downtime during maintenance windows.
Exactly! That's the reason why proper maintenance windows are scheduled for resources.