Upgrade Grafana Alloy from 0.12.6 to 1.13+

Summary

Upgrade the default Grafana Alloy Helm chart version from 0.12.6 to 1.13+ (current stable: v1.13.0).

Why this is needed

Chart 0.12.6 is a very old pre-1.0 release of Grafana Alloy (the successor to Grafana Agent). The 0.x series predates the stable 1.0 release and is no longer supported. Running a pre-1.0 observability agent on Kubernetes 1.36 clusters is untested and potentially incompatible with current Grafana ecosystem integrations (Loki, Mimir, Prometheus remote write).

This upgrade is a prerequisite for the planned Kubernetes 1.36 upgrade.

Relevant links:

• Grafana Alloy v1.13.0 release notes: https://github.com/grafana/alloy/releases/tag/v1.13.0
• Grafana Alloy Kubernetes install docs: https://grafana.com/docs/alloy/latest/set-up/install/kubernetes/
• Grafana Alloy Helm chart: https://grafana.github.io/helm-charts

Backward compatibility

Grafana Alloy v1.13.x does not publish an explicit minimum Kubernetes version in its Helm chart kubeVersion constraint. Based on the chart templates, it is compatible with Kubernetes 1.22+ at minimum. It can be deployed in place on clusters currently running K8s 1.31 through 1.35 — this upgrade can be performed independently of and prior to the Kubernetes 1.36 upgrade.

Recommend validating on a staging cluster before rolling broadly, as the 0.x → 1.x jump involves a significant feature surface change.

Breaking changes in the upgrade

The 0.x → 1.x line introduced configuration changes to Alloy's River/Alloy config syntax. Review any custom Alloy configuration blocks for deprecated or renamed components. See the Alloy migration docs for specifics.

Acceptance criteria

• Default Grafana Alloy chart version updated to 1.13+ in workload.go (AWS and Azure defaults)
• Alloy config syntax reviewed for 0.x → 1.x breaking changes
• Verified metrics/logs pipeline intact on a staging cluster after upgrade

Start in workload.go and locate the default Grafana Alloy Helm chart versions for the AWS and Azure workloads. Review the Alloy migration and release notes for 0.x to 1.x configuration changes, then validate the updated chart and configuration on a staging Kubernetes cluster. Done means both defaults use 1.13+ and the metrics and logs pipelines remain intact.