We could have a ton of these various timeout issues in Tempest, we're never going to nail these down properly so we need to step back and start looking at Tempest and thinking about how to make that smarter than assuming a hard-coded value is going to fit all scenarios, i.e. there could be concurrent load on the host causing it to run slow at the time of failure, or we could look at getting diagnostic information back about the instance and logging that to see if there is something more specific to fingerprint when we hit a timeout failure.
We could have a ton of these various timeout issues in Tempest, we're never going to nail these down properly so we need to step back and start looking at Tempest and thinking about how to make that smarter than assuming a hard-coded value is going to fit all scenarios, i.e. there could be concurrent load on the host causing it to run slow at the time of failure, or we could look at getting diagnostic information back about the instance and logging that to see if there is something more specific to fingerprint when we hit a timeout failure.