Skip to main content

Cluster Operators

This section is for cluster operators — people who run and maintain OLIVE in production or lab environments: Kubernetes, networking, storage, ingress, observability, backups, and the supporting stack described in the platform technology overview.

Use it when you need operational context: what components exist, how they fit together, and what to consider when operating or troubleshooting the platform.

This guide does not define the cluster operator role as a job. It assumes you already run Kubernetes — workloads, RBAC, ingress, storage and the cluster's own tooling — and documents what OLIVE needs from it, which is the part you cannot see from outside.

Two things are worth knowing up front. The components OLIVE depends on are installed with cluster rights, which is why Infrastructure Charts lives in this guide rather than the Platform Administrator's. And who owns which control is set out in Shared Responsibility.

Start here if an OLIVE deployment is coming to your cluster

Handover Checklist is the single list the cluster operator and the platform administrator work through together before the first install: what you provide, the values each side needs from the other, the mesh exemptions, and the cluster-scoped checks only you can run. Most of the values exchanged there fail silently when they are wrong.

Getting Started​

  • Handover Checklist — the pre-install handover between you and the platform administrator, in the order it happens
  • Platform Technology Overview — concise map of main technologies, versions, and roles (Kubernetes, Keycloak, Vault, databases, storage, ingress, and supporting services)
  • Kubernetes Resources — which API groups and CRDs OLIVE needs, what it creates cluster-wide, and an Argo CD AppProject allow-list

Core Operations​

Security​

Infrastructure​

  • Infrastructure Charts — Vault, Keycloak, the Istio mesh configuration and Kyverno: what OLIVE needs of each, and the optional as-is charts that set them up
  • Networking — DNS, cluster networking, egress carve-outs, firewall rules, and the default-deny network policies
  • Storage Requirements — what OLIVE claims, which access modes it needs, and how the StorageClass is chosen
  • Ingress Requirements — the hostnames per tenant, what the ingress controller must support, and the TLS Secret OLIVE expects
  • What to Back Up — where the state lives, which pieces must be consistent with each other, and the order a restore has to follow

Troubleshooting​

  • Troubleshooting — the common OLIVE failure modes by symptom; most of them are silent, and several are cluster-side rather than configuration

Reference​


Note: For context on where OLIVE is deployed and how tenants are isolated, see Deployment Options and Tenant Isolation in the Platform Administrators section.