Skip to content

SLA & Shared Responsibility

Trust is built on transparency. Cyberun Cloud adheres to the industry-standard Shared Responsibility Model and provides clear Service Level Agreements (SLA) to ensure your critical business operations have predictable reliability.

Shared Responsibility Model

Similar to AWS or Azure, we divide security and operational responsibilities into "Security OF the Cloud" and "Security IN the Cloud."

Layer Responsible Party Definition
Data Customer Application-level encryption, access control (IAM user management), data classification.
Application Customer Code deployed in containers, library dependency updates, application-layer firewall rules.
Container Platform Cyberun Kubernetes Control Plane (API/Etcd), Cilium Network, Karmada Federation, Registry Security.
Infrastructure Cyberun Physical servers, OS hardening, WireGuard Mesh, Storage Cluster health.
Physical Cyberun Power, cooling, and physical security (provided by our Tier-III+ data center partners).

Service Level Objectives (SLO) & Agreements (SLA)

Based on our "Production Ready" architectural standards, we offer the following availability commitments.

Compute & Control Plane

  • SLA: 99.95% (Monthly)
  • Definition: The API Server is accessible and capable of scheduling Pods.
  • Guarantee Mechanism: 3-Replica HA deployment in the Tokyo control plane and cross-AZ redundancy.

Storage Durability

  • SLA: 99.99% (Availability) | 99.99999% (Durability)
  • Definition: Written data is readable and not lost.
  • Guarantee Mechanism: Ceph 3-Replica strong consistency replication and background data scrubbing.

Network Connectivity

  • SLA: 99.9% (Cross-Region Connectivity)
  • Definition: Packet loss rate within the inter-region WireGuard mesh is < 1%.
  • Guarantee Mechanism: BGP automatic route failover and multi-link redundancy.

Maintenance Windows

To maintain platform security and performance, we perform regular infrastructure updates.

  • Zero-Downtime Maintenance: For control plane upgrades and node kernel patching, we utilize Rolling Update strategies to ensure zero business impact.
  • Notifications: For major network changes that may impact performance, we will notify users via the Status Page at least 48 hours in advance.