Monitoring AWS Resources

Introduced in GitLab 9.4

GitLab has support for automatically detecting and monitoring AWS resources, starting with the Elastic Load Balancer. This is provided by leveraging the official Cloudwatch exporter, which translates Cloudwatch metrics into a Prometheus readable form.

Requirements

The Prometheus service must be enabled.

Metrics supported

Name Query
Throughput (req/sec) sum(aws_elb_request_count_sum{%{environment_filter}}) / 60
Latency (ms) avg(aws_elb_latency_average{%{environment_filter}}) * 1000
HTTP Error Rate (%) sum(aws_elb_httpcode_backend_5_xx_sum{%{environment_filter}}) / sum(aws_elb_request_count_sum{%{environment_filter}})

Configuring Prometheus to monitor for Cloudwatch metrics

To get started with Cloudwatch monitoring, you should install and configure the Cloudwatch exporter which retrieves and parses the specified Cloudwatch metrics and translates them into a Prometheus monitoring endpoint.

Right now, the only AWS resource supported is the Elastic Load Balancer, whose Cloudwatch metrics can be found here.

A sample Cloudwatch Exporter configuration file, configured for basic AWS ELB monitoring, is available for download.

Specifying the Environment label

In order to isolate and only display relevant metrics for a given environment however, GitLab needs a method to detect which labels are associated. To do this, GitLab will look for an environment label.