This directory contains AMP-specific replacements for files in examples/dora/. Use these instead of the parent directory files when your Prometheus backend is AMP.

Architecture

GitHub Actions  ──push──►  Pushgateway (in-cluster)  ──scrape──►  Prometheus Agent
                                                                         │
                                                              remote_write (SigV4/IRSA)
                                                                         │
                                                                         ▼
                                                              Amazon Managed Prometheus
                                                                         │
                                                                    Grafana / AMG

AMP has no public Pushgateway endpoint. The supported pattern is:

GitHub Actions pushes deploy/incident metrics to an in-cluster Pushgateway (same as the default variant — no change to workflow steps)
A Prometheus Agent scrapes the Pushgateway and remote_writes to AMP using SigV4 (via IRSA)
Recording rules are managed via Terraform (preferred) or the AWS CLI fallback script
Grafana connects to AMP via SigV4 (self-hosted with plugin, or Amazon Managed Grafana)

Files

File	What it does
`amp-workspace.tf`	Terraform: provision AMP workspace + deploy DORA recording rules in one apply
`pushgateway-helm-values.yaml`	Prometheus Pushgateway Helm values (in-cluster)
`prometheus-agent-values.yaml`	Prometheus Agent Helm values: scrapes Pushgateway, remote_writes to AMP via SigV4
`amp-recording-rules-deploy.sh`	AWS CLI fallback: create/update DORA rules in AMP without Terraform
`grafana-amp-datasource.yaml`	Grafana datasource ConfigMap — self-hosted Grafana → AMP (SigV4, IRSA)
`grafana-amg-datasource.json`	Datasource config for Amazon Managed Grafana (SigV4 auto-handled)

Files from examples/dora/ that are unchanged for AMP:

deployment-event-step.yaml — still pushes to in-cluster Pushgateway, no changes
incident-webhook-handler.yaml — still pushes to in-cluster Pushgateway, no changes
prometheus-recording-rules.yaml — same rule YAML; amp-workspace.tf and amp-recording-rules-deploy.sh both consume it
grafana-dashboard.json — same dashboard JSON; import into AMG or self-hosted Grafana pointing at AMP

Quick start — Terraform path (recommended)

# 1. Provision AMP workspace + deploy DORA recording rules
cd examples/dora/amp-variant
terraform init
terraform apply -var="workspace_alias=dora-platform"

# Capture the outputs for steps 2 and 4:
terraform output workspace_prometheus_endpoint
terraform output workspace_id

# 2. Deploy in-cluster Pushgateway
helm upgrade --install pushgateway prometheus-community/prometheus-pushgateway \
  --namespace monitoring --create-namespace \
  -f pushgateway-helm-values.yaml

# 3. Deploy Prometheus Agent
# Replace <endpoint> with the remote_write_url Terraform output.
# Replace <account-id> with your AWS account ID in prometheus-agent-values.yaml first.
helm upgrade --install prometheus-agent prometheus-community/prometheus \
  --namespace monitoring \
  -f prometheus-agent-values.yaml

# 4. Connect Grafana to AMP
kubectl apply -f grafana-amp-datasource.yaml
# Then import examples/dora/grafana-dashboard.json into Grafana

Quick start — AWS CLI fallback (no Terraform)

# 1. Create AMP workspace
aws amp create-workspace --alias dora-platform --region eu-central-1

# 2. Deploy DORA recording rules
AMP_WORKSPACE_ID=$(aws amp list-workspaces \
  --query 'workspaces[?alias==`dora-platform`].workspaceId' \
  --output text --region eu-central-1)
bash examples/dora/amp-variant/amp-recording-rules-deploy.sh "$AMP_WORKSPACE_ID"

# 3-4. Same as Terraform path above (Helm + Grafana)

Prerequisites

Component	Requirement
Terraform	>= 1.5.7 (for `amp-workspace.tf`)
AWS provider	>= 6.28
IRSA / Pod Identity	Prometheus Agent pod needs `AmazonPrometheusRemoteWriteAccess`
IRSA / Pod Identity	Grafana pod needs `AmazonPrometheusQueryAccess`
Helm	>= 3.x
yq	v4+ (for dora-validate.sh)