問題文
An engineer wants a chart that shows the mean CPU utilization of the hosts running a particular service, with one line per data center. Which sequence of actions on a single plot produces that?
選択肢
- Enter the metric as the signal, add the mean function as a transformation, and set a moving window of one hour, because transformations are what produce one output series per group.
- Create one plot per data center, each with its own filter for that data center, and apply the mean function to each; grouping is not available on a single plot so this is the only way to get one line per data center.
- Enter the metric as the signal, add a filter for the service dimension, then add the mean function as an aggregation and set the group-by to the data center dimension.
- Enter the metric as the signal and rely on the default behavior, which already averages across the matching hosts and splits the result by the data center dimension on its own.