Volume Monitoring in Kubernetes with PrometheusArgo workflows create pods and deployments on Kubernetes. When running on AWS, we use EBS for volumes. AWS provides some basic volume metrics but only down to a 5 minute granularity and does not provide any filesystem-level stats, making it hard, for example, to tell how much of the volume capacity was actually being used. To address these issues, we decided to add our own monitoring usingiostats
