To date, we have only ever processed ndt datatypes so metrics with only a datatype label were sufficient. Or, owing to historical datatypes, synonymous with the experiment, e.g. paris-traceroute. As we consider adding additional experiments with the same sidecar datatypes as ndt, we need additional labels to clearly distinguish processing for ndt or another experiment.
These metrics from the Pipeline: Overview dashboard include:
query_cost_seconds_sum
gardener_bytes_sum - adjust label value (strip "-json" suffix from experiment label)
To date, we have only ever processed
ndtdatatypes so metrics with only a datatype label were sufficient. Or, owing to historical datatypes, synonymous with the experiment, e.g. paris-traceroute. As we consider adding additional experiments with the same sidecar datatypes as ndt, we need additional labels to clearly distinguish processing for ndt or another experiment.These metrics from the
Pipeline: Overviewdashboard include: