Managed Kubernetes
Concepts
Entdecken Sie die Grundlagen und wesentlichen Prinzipien, um unsere Infrastruktur sicher zu beherrschen.
Konzepte erkunden →Tutorials
Lernen Sie Schritt für Schritt, wie Sie unsere Dienste mit detaillierten Anleitungen konfigurieren und nutzen.
Tutorials entdecken →Managed Kubernetes by Cloud Temple ist eine Container-Orchestrierungslösung auf Basis von Open-Source-Produkten und wurde speziell entwickelt, um ein hohes Maß an Sicherheit, Resilienz und Automatisierung auf den SecNumcloud-Plattformen von Cloud Temple zu gewährleisten. Jeder Cluster wird in einer vollständig dedizierten IaaS-Umgebung von Cloud-Temple OpenIaaS für den Kunden bereitgestellt.
Diese Lösung ist für Teams konzipiert, die über fundierte Kenntnisse von Kubernetes verfügen und Anhänger des Open Source-Paradigmas sind, die eine native, portierbare Lösung ohne Hersteller-spezifische Zusatzschichten suchen, die auf einem minimalen und unveränderlichen Betriebssystem basiert, das speziell für Automatisierung und Sicherheit entwickelt wurde.
Key Benefits
- Sovereignty and Reversibility: The solution is built on open-source standards (CNCF Kubernetes) to avoid any technological lock-in and ensure application portability. The Veeam Kasten backup tool included in the offering is specifically designed to simplify migrations between clouds.
- Zero-Trust Security and Governance: The architecture is based on Talos OS, an immutable operating system with no direct access (no shell, no SSH), drastically reducing the attack surface. This approach is combined with governance tools such as Kyverno for policy management and Capsule for fine-grained access control, forming a solid foundation for a Zero-Trust security strategy.
- Cost Control and Integration: The solution natively integrates FinOps tools like OpenCost for precise consumption tracking. The economic model is transparent, based on consumed IaaS resources, and the use of well-established open-source components (Cilium, Ceph, ArgoCD) facilitates seamless integration into your existing ecosystems.
A complete and ready-to-use platform
The solution natively includes a comprehensive and consistent stack of cutting-edge open-source tools to cover all requirements throughout the application lifecycle:
- Networking and Security: Cilium, Hubble, MetalLB, Ingress Nginx, Kyverno, Capsule
- Storage: Rook-Ceph
- Continuous Deployment (GitOps): ArgoCD
- Observability: Prometheus, Grafana, Loki
- Backup and Migration: Veeam Kasten
- Cost Management (FinOps): OpenCost
Deployment Architectures
We offer two distinct architectures to meet your needs, whether for development environments or critical production setups.
Architecture "Dev/Test"
Ideal für POC-Umgebungen, diese kompakte Architektur stellt alle Ressourcen innerhalb einer einzigen Verfügbarkeitszone (AZ) bereit.
- Verwendungszweck: Entwicklung, Tests, Proof-of-Concept (POC).
- Wichtige Merkmale:
- 1 Control-Plane-Knoten.
- 3 Worker-Knoten (oder mehr).
- Der verteilte Speicher (Ceph) ist auf den Worker-Knoten ko-lokalisiert.
- Kein SLA für hohe Verfügbarkeit.
- Keine Sicherheitsbeschränkungen.
Production Architecture (Multi-AZ)
Designed for production and critical applications, this architecture distributes resources across three Availability Zones (AZ) to ensure high availability and maximum resilience, in compliance with SecNumCloud requirements.
- Use Case: Production applications, critical services, platforms requiring an SLA.
- Key Features:
- High Availability: 3 Control Plane nodes distributed across 3 AZs.
- Dedicated Storage: 3 dedicated storage nodes distributed for performance and resilience.
- Distributed Workers: Minimum of 3 worker nodes, one per AZ.
- Bare Metal Nodes (Optional): Optional integration of "Bare Metal" worker nodes for specific performance requirements, particularly GPU support.
- SLA of 99.90%.
Detailed Technical Components
The offering includes the following components in detail:
- CNI Cilium, with observability interface (Hubble)
- Internal and external ingresses via MetalLB and nginx
- Distributed storage with Rook-Ceph
- Cert-Manager
- ArgoCD
- Prometheus stack (Prometheus, Grafana, Loki)
- Container registry Harbor
- Cost management with OpenCost
- Advanced security policies with Kyverno and Capsule
- Veeam Kasten (backup, cross-environment automation, and reversibility)
- SSO authentication with external Identity Provider OIDC (Microsoft Entra, FranceConnect, Okta, AWS IAM, Google, Salesforce, ...)