What lightweight monitoring option can I run that will give me historical stats on memory and cpu for both the host and each container? I'm trying to track down a random timeout (suspecting caused by cpu spikes) at exactly 2am on some nights.
Should run as a docker container. Thanks!