When enabling zone redundancy for an on-demand (default of zero ready-instances) Azure Function, the action effectively converts the function to an always-ready configuration with two instances; this was described in TrackingID#2603270050000503.
This documentation was quoted in that ticket:
https://learn.microsoft.com/azure/reliability/reliability-functions?pivots=flex-consumption#instance-distribution-across-zones
https://github.com/MicrosoftDocs/reliability-docs/blob/main/articles/reliability/reliability-functions.md
My reading of the current documentation is that zone redundancy for on-demand should not create always-ready instances, but I was told this is incorrect.
The documentation needs improvement, maybe adding a clearer warning that 'scale to zero' (ie. serverless) for on-demand and zone redundancy is incompatible and by checking the box it sets always-ready instances to two.
When enabling zone redundancy for an on-demand (default of zero ready-instances) Azure Function, the action effectively converts the function to an always-ready configuration with two instances; this was described in TrackingID#2603270050000503.
This documentation was quoted in that ticket:
https://learn.microsoft.com/azure/reliability/reliability-functions?pivots=flex-consumption#instance-distribution-across-zones
https://github.com/MicrosoftDocs/reliability-docs/blob/main/articles/reliability/reliability-functions.md
My reading of the current documentation is that zone redundancy for on-demand should not create always-ready instances, but I was told this is incorrect.
The documentation needs improvement, maybe adding a clearer warning that 'scale to zero' (ie. serverless) for on-demand and zone redundancy is incompatible and by checking the box it sets always-ready instances to two.