linux-monitoring-cloudwatch

Send MemoryUsage percentage metrics to AWS CloudWatch from Linux-based EC2 instances.

Why is this useful?

EC2 natively only sends hypervisor-visible metrics to CloudWatch (e.g. CPU usage). Others such as memory usage require OS-level integration - this is where linux-monitoring-cloudwatch comes in.

No further dependencies are necessary on your instance - just the executable.

With the metrics emitted by linux-monitoring-cloudwatch (MemoryUsage at this time), you may trigger alarms & autoscaling.

Example usage

./linux-monitoring-cloudwatch --namespace my-app/int --region eu-west-1

--region is optional: without providing it, linux-monitoring-cloudwatch will query the instance metadata API at runtime.

Building (with Docker)

docker run -v "$PWD":/go/src/github.com/darnould/linux-monitoring-cloudwatch -w /go/src/github.com/darnould/linux-monitoring-cloudwatch golang:1.3 go build

This will produce a single linux-monitoring-cloudwatch executable.

Installation

• Build or download the executable.
• Copy the linux-monitoring-cloudwatch executable to your instances' filesystems.
• Allow your instances to make PutMetricData calls to CloudWatch with an IAM policy such as:

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Effect": "Allow",
            "Action": [
                "cloudwatch:PutMetricData",
            ],
            "Resource": [
                "*"
            ]
        }
    ]
}

• A cron job such as the following will report metrics to CloudWatch every 5 minutes:

/5 * * * *  /path/to/linux-monitoring-cloudwatch --namespace NAMESPACE --region REGION &>> /var/log/linux-monitoring-cloudwatch.log

Coming up

• DiskUsage percentage metric