Before reading about the EBS default policy, it is important to understand the following Metricly computed metrics.
|Policy name||Duration||Condition 1||(and) Condition 2||Cat.||Description|
|Elevated Queue Length Differential with Elevated Latency||30 min||metricly.aws.ebs.queuelengthdifferentialhas an upper baseline deviation + static threshold > 1||metricly.aws.ebs.averagelatency has an upper baseline||CRITICAL||The first condition of the policy looks for an upper deviation as the first indication that the disk may be getting more traffic than it can keep up with. It also checks for the differential to be greater than 1 in order to avoid false alarming in cases where the differential is very low.The second condition is added because an elevated queue differential by itself is not necessarily a bad thing. We only want to alarm if your differential is higher than normal AND your latency is higher than normal.|
|Elevated iops Utilization with low Burst Balance||5 min||netuitive.aws.ebs.iopsutilization => 90||aws.ebs.burstbalance <= 10.||This policy looks at two metrics: IOPS Utilizaton and EBS Burst Balance. High IOPS Utilization (a Metricly computed metric) indicates the disk is highly utilized and a low EBS Burst Balance indicates the disk is so highly utilized that most of the burst available to the disk is depleted. Once the burst balance is fully depleted available disk IOPS will fall causing slowdowns in I/O and deteriorated performance of the application using the volume.\n\nCheck this volume and the application using it to see if the I/O profile has changed. Consider using Provisioned IOPS to increase the disk performance if this new profile is normal.|