Downtime for updated images

[paynecodes]

We had some downtime in our App Service today that appears to have been caused by container images being deployed (based on timing of log messages and know downtime window).

Is downtime expected? I was under the impression that a new container should be deployed and accepting traffic before the old container is taken offline.

If this is expected / likely to continually reoccur:
What are the best strategies to mitigate and/or speed up deployment to improve long downtimes such as these (16 minutes in this case).

If this is not expected or likely to have been the issue:

What logging options and configuration changes are necessary to be able to get more from logs via Log Stream?

[samrueby]

App Service indeed wait until that new container is running and ready for HTTP requests before they switch over to it. 16 minutes is quite a long time- that is unexpected. The container starts for me < 2 minutes.

I believe the best logging option would be to configure a 'diagnostic setting' for your app service. Perhaps App Service Console Logs, App Service Application Logs, App Service Platform logs. Send to log analytics.