aws-rds-engine-version-prometheus-exporter

Prometheus Exporter for AWS RDS Engine Version

How to run

Local

$ go run main.go

Binary

Get the binary file from Releases and run it.

Docker

$ docker run chaspy/aws-rds-engine-version-prometheus-exporter:v0.1.0

Metrics

$ curl -s localhost:8080/metrics | grep aws_custom_rds_cluster_count
# HELP aws_custom_rds_cluster_count Number of RDS
# TYPE aws_custom_rds_cluster_count gauge
aws_custom_rds_cluster_count{cluster_identifier="api-postgres-develop-a",engine="aurora-postgresql",engine_version="10.7",eol_status="ok"} 1
aws_custom_rds_cluster_count{cluster_identifier="api-postgres-production-a",engine="aurora-postgresql",engine_version="10.7",eol_status="ok"} 1
aws_custom_rds_cluster_count{cluster_identifier="video-production-a",engine="aurora-postgresql",engine_version="9.6.17",eol_status="ok"} 1
aws_custom_rds_cluster_count{cluster_identifier="video-staging-a",engine="aurora-postgresql",engine_version="9.6.17",eol_status="ok"} 1

metric	description	tags	note
aws_custom_rds_eol_status_ok	Number of instances whose EOL status is ok	"cluster_identifier", "engine", "engine_version"
aws_custom_rds_eol_status_alert	Number of instances whose EOL status is alert	"cluster_identifier", "engine", "engine_version"
aws_custom_rds_eol_status_warning	Number of instances whose EOL status is warning	"cluster_identifier", "engine", "engine_version"
aws_custom_rds_eol_status_expired	Number of instances whose EOL status is expired	"cluster_identifier", "engine", "engine_version"
aws_custom_rds_cluster_count	Number of RDS	"cluster_identifier", "engine", "engine_version", "eol_status"	DEPRECATED

IAM Role

The following policy must be attached to the AWS role to be executed.

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "VisualEditor0",
            "Effect": "Allow",
            "Action": [
                "rds:DescribeDBInstances",
                "rds:DescribeDBClusters"
            ],
            "Resource": "*"
        }
    ]
}

Environment Variable

name	required	default	description
ALERT_HOURS	no	2160	Time to determine "alert" status for EOL dates
WARNING_HOURS	no	4320	Time to determine "warning" status for EOL dates
AWS_API_INTERVAL	no	300	Interval between calls to the AWS API

Datadog Autodiscovery

If you use Datadog, you can use Kubernetes Integration Autodiscovery feature.