We have some tests with some variability that can't currently be well handled by the existing metadata system - for example runtime metrics. In runtime-metrics we are currently trying to separate the telemetry by runtime variation but using the same mechanism (metadataConfig) as we do for configuration-based variations.
On top of this, we aren't running exhaustive tests for the runtime variations, and are sometimes seeing some churn in the emitted attributes.
See:
We should determine if we need some additional when capabilities that can support things like runtime variations on top of standard config variations.
We have some tests with some variability that can't currently be well handled by the existing metadata system - for example runtime metrics. In runtime-metrics we are currently trying to separate the telemetry by runtime variation but using the same mechanism (
metadataConfig) as we do for configuration-based variations.On top of this, we aren't running exhaustive tests for the runtime variations, and are sometimes seeing some churn in the emitted attributes.
See:
We should determine if we need some additional
whencapabilities that can support things likeruntimevariations on top of standardconfigvariations.