Use cases

This topic provides information about some of the key metrics to monitor the performance of the EC2 system.

Monitoring the availability of the EC2 instances

PATROL for Amazon EC2 KM discovers and monitors the EC2 instances. An AWS_EC2 application instance is created for each Amazon EC2 instance. An AWS_EC2 instance is sometimes attached to and appear under an AWS_ELB instance, or it can appear under the AWS_ACCOUNT, when not attached to an ELB instance. You can monitor the _InstanceStatus parameter of AWS_EC2 application class to monitor its availability. This parameter will indicate whether the EC2 instance is in the running, stopped or terminated state.

InstanceStatus parameter reports following values:

0= Running, pending

1=Stopped

2=Terminated

The KM generates a WARNING whenever an EC2 instance is stopped and an ALARM alert when EC2 is terminated. You can configure (or modify) these thresholds as per your need.

Monitoring the health of the EC2 instances

PATROL for Amazon EC2 KM discovers and monitors the EC2 instances. An AWS_EC2 application instance is created for each Amazon EC2 instance. An AWS_EC2 instance is sometimes attached to and appear under an AWS_ELB instance, or it can appear under the AWS_ACCOUNT, when not attached to an ELB instance. You can monitor the following metrics to assess the health of EC2 instance:

• CPUUtilization - It displays the percentage of allocated EC2 computer processing units that are currently in use on the instance.
• DiskWriteOps - It displays the number of completed write operations from all hard disks that are available to the EC2 instance.
• DiskReadOps - It displays the number of completed read operations from all hard disks available to the EC2 instance.
• NetworkOut - It displays the number of bytes received on all network interfaces by the EC2 instance.
• NetworkIn - It displays the number of bytes received on all network interfaces by the EC2 instance.

You can configure the appropriate thresholds on these parameters to monitor the increase in network I/O or disk operations. You can also configure the alert on CPU utilization of EC2 instance if it goes above acceptable limits.

For example: >75%