Extra telemetry on policy evaluation failure

[the-nando]

We recently run into two separate issue where the Nomad autoscaler failed to describe AWS autoscaling groups due to an expired AWS token or failed to evaluate a scaling policy because of an issue reaching the APM (Prometheus).

{"@level":"warn","@message":"failed to get target status","@module":"policy_manager.policy_handler","@timestamp":"2023-07-06T16:21:26.029652Z","error":"failed to describe AWS Autoscaling Group: operation error Auto Scaling: DescribeAutoScalingGroups, https response error StatusCode: 403, RequestID: c674bc86-1234-4fb1-5678-b264741176bc, api error ExpiredToken: The security token included in the request is expired","policy_id":"613aeb80-xs23-8f4e-1234-ef2ca2748d8a"}

It would be great to have a couple of extra Prometheus metrics exported by the autoscaler to be monitored to detect simple failures.