I am trying to get a ecs dashboard set up in datadog including the number of ecs tasks running over the past timeframe.
avg:ecs.containerinsights.running_task_count.sum{$ecs_service,$ecs_cluster} by {servicename}.as_count()
This works correctly over 1 or 4 hours, but over longer periods it ends up with larger numbers. The longer, the larger so it seems to be summing everything up.
The answer is to use the rollup() function.
https://docs.datadoghq.com/dashboards/functions/rollup/