Closed
Description
Description
When running cortex cluster up or cortex cluster configure, the CLI should check that the user's AWS account has enough vCPU quota to support max_instances instances of the chosen instance type.
Question
What would be the strategy for spot instances (or would this check only be applied when not using spot instances)? With spot, there could be multiple instance types, and the spot quota is a different quota than for on-demand (see here).