Related: T371321: [Idea] Collect pageview data using client-side instrumentation. I'd expect that energy consumption would go down if we stopped searching for needles in the webrequest haystack.

FWIW. Almost all of the power consumption in analytics infra (understandably) is coming from an-worker nodes. There is some large consumption coming from presto cluster too: https://grafana.wikimedia.org/goto/ytdntxGDR?orgId=1

Since the vast majority of analytic jobs are scheduled via Airflow, we could certainly change the schedule cron definitions of each DAG to achieve this, but it would require careful planning, as today we rely heavily on the @daily alias, which makes everything run at midnight UTC.

We could certainly pick one heavy job and do an experimental rescheduling as a start.

In T384166#11565913, @xcollazo wrote:

Since the vast majority of analytic jobs are scheduled via Airflow, we could certainly change the schedule cron definitions of each DAG to achieve this, but it would require careful planning, as today we rely heavily on the @daily alias, which makes everything run at midnight UTC.

That seems to be 7pm virgina time. If we push the default a bit earlier, that could automatically have the intended effect.

We could certainly pick one heavy job and do an experimental rescheduling as a start.

That sounds good to me!