Skip to content

Latest commit

 

History

History
79 lines (56 loc) · 5.85 KB

extension-roadmap.md

File metadata and controls

79 lines (56 loc) · 5.85 KB

Extension Roadmap

All ArgoPlane extensions are named by domain, not by the underlying tool. Extensions focus on Observe (workload visibility) and Secure (security and compliance) categories.

Strategy

ArgoPlane is a collection of ArgoCD UI extensions. Each extension attaches operational data to resources and applications inside ArgoCD: resource tabs, app views, and occasionally a system-level sidebar page for cross-cutting dashboards. Every extension works standalone and has no knowledge of the others.

ArgoCD Extension Points

ArgoCD v3.x provides several registration methods. ArgoPlane uses these:

Method Where it appears Used by
registerResourceExtension Tab on a K8s resource's sliding panel Metrics, Networking, Logs, Vulnerabilities, Events
registerAppViewExtension Full-page view in app details Metrics, Networking, Logs, Vulnerabilities, Events
registerSystemLevelExtension Sidebar page (global, not tied to any app) ArgoPlane Overview (planned)
registerTopBarActionMenuExt Action button in top toolbar (reserved for future use)
Application tab (registerResourceExtension with argoproj.io/Application) Tab on Application resource (reserved for future use)

Done

Category Domain Description Backend Extension Points
Observe metrics CPU, memory, request rates, latency, custom PromQL queries Prometheus Resource tabs (Deployment, StatefulSet, Pod) + App view
Observe networking Traffic flows (Hubble), network policies, app vs platform ownership Cilium/Hubble App view + Resource tabs (Pod, CiliumNetworkPolicy, CiliumClusterwideNetworkPolicy)
Observe logs Historical log search, label-based filtering, severity detection, time range selection, volume charts. No live tail (ArgoCD's built-in Logs tab handles real-time pod streaming; Loki's tail API uses WebSocket which the proxy extension mechanism does not support). Loki Resource tabs (Pod, Deployment, StatefulSet) + App view
Secure vulnerabilities Image vulnerability scanning (CVEs), config audit, exposed secrets detection, SBOM inventory. Per-image severity breakdown, CSV export, deduplication across ReplicaSets. Trivy Operator (K8s API) App view (4 tabs: Vulnerabilities, Config Audit, Exposed Secrets, SBOM)
Observe events Kubernetes events per resource and application. Warning/Normal event filtering, timeline view. Kubernetes API Resource tabs + App view
Observe alerts Firing and pending alerts per app, app vs platform ownership from the PrometheusRule that defines them, rule health, runbook and dashboard links. Read-only; silences planned. Prometheus rules API (Prometheus, Thanos, Mimir, vmalert), optional Alertmanager App view + Resource tabs (Pod, Deployment, StatefulSet, DaemonSet, Job)

Next

Category Domain Description Backend Extension Points
Secure policies Policy violations, admission control results, policy reports per app Kyverno Resource tabs + App view

System-Level Extensions (Planned)

Feature Extension Point Description
ArgoPlane Overview registerSystemLevelExtension Sidebar page with aggregated health: all apps' sync status, top alerts, cluster CPU/memory. Platform engineer's dashboard inside ArgoCD.
Alerts Dashboard registerSystemLevelExtension All firing/pending alerts across all apps. Filter by team, namespace, severity. Global view of what's wrong.

Future

Category Domain Description Backend
Observe traces Distributed traces, spans, latency breakdown, trace-to-log correlation Tempo, Jaeger
Secure certificates TLS certificate expiry warnings, renewal status, issuer details cert-manager
Observe scaling HPA/VPA status, scaling events, recommendations, KEDA triggers Kubernetes HPA, KEDA, VPA

Not Planned

Domain Reason
notifications ArgoCD already has argocd-notifications
identity ArgoCD native RBAC is sufficient
gitops ArgoCD itself handles this
catalog / self-service / tenant management Out of scope. ArgoPlane is an extension package, not a developer portal.

Maybe Later

Category Domain Description Notes
Observe databases DB health, connections, backups, scaling CNPG, Percona, Redis Operator
Observe costs Resource cost attribution per app/namespace Kubecost, OpenCost
Secure secrets Secret sync status, rotation, store health External Secrets, Vault; ESO already has visibility tools
Observe builds Build status, image tags, pipeline runs GitHub Actions, GitLab CI, Tekton

Design Principles

Domain-specific APIs. Each extension backend abstracts the underlying tool behind a domain-specific API. Swapping Prometheus for OTel or Loki for Elasticsearch should only require changing the backend implementation, not the UI or API contract.

Contextual relevance. Surface information where it matters. A Pod tab should show that pod's logs and alerts, not a generic search page. App views provide the broader picture.

App vs platform ownership. Where applicable (alerts, policies, networking), distinguish between resources the app team deployed and platform-level resources that affect them. This pattern is established in the networking extension and should be reused.

Smart links. Connect the dots between systems. An alert should link to its Grafana dashboard, runbook, and source YAML. A policy violation should link to the policy definition. Logs should link to the trace ID.