Skip to content

Latest commit

 

History

History
80 lines (71 loc) · 2.74 KB

pipeline_into_the_details.md

File metadata and controls

80 lines (71 loc) · 2.74 KB

Pipeline into the details

‘Prerequisite’ / ‘Deployment’ / ‘Configuration’ - Management cluster

Using Assisted Installer


--- PRIVATE AND CONFIDENTIAL, DO NOT DISTRIBUTE ---


  1. Prerequisite Stage:
    1. Manual Steps:
      • BIOS & Firmware updates on all servers
      • Networks configurations:
        • Switch configuration
        • Subnet configurations
    2. Ansible Orchestration - via jumphost:
      • Generate SSH key pair used for debugging deployments
      • Set up local container for the registry
      • The Container registry for hosting and distributing additional container images:
        • openshift-release-dev/ocp-release
        • ocpmetal/assisted-installer-controller
        • ocpmetal/assisted-installer-agent
        • ocpmetal/assisted-installer
        • olm-index/redhat-operator-index
      • HTTP Store (optional):
        • Is HTTP Store a pre-exisiting server?
        • Setup HTTP Server
      • DNS (optional):
        • Is DNS pre-existing?
        • Setup DNSMASQ DNS Server
      • NTP (optional):
        • Is NTP pre-existing?
        • Setup Chrony NTP Server
      • Podman creates the Assisted Installer pod:
        • Installer
        • DB
        • UI
      • Launch stand alone Assisted Installer with REST API capabilities.
  2. Configuration:
    • Assisted Installer:
      • Create cluster deployment
      • Configure cluster deployment
        • Redirects for the repose of the registry:
          • QUAY
          • SSL
      • Restricted network install customizations (optional)
      • Network SDN drivers (optional):
        • Openshift SDN
        • OVN Kubernetes
      • API discovery ISO customizations:
        • Static IPs:
      • Download the Discovery ISO:
        • Copy to the HTTP Store.
  3. Discovery:
    • Mount the Discovery ISO from the HTTP Store on target hosts
    • Set ISO to boot media on target hosts
    • Reboot hosts into Discovery ISO
  4. Post discovery:
    • Apply additional cluster install configurations:
      • Ingress VIP
      • API Controller
    • Assign roles to discovered hosts
  5. Deploy / Install Management Cluster:
    • Trigger install via the Assisted Installer REST API.
    • Monitor install via the Assisted Installer REST API.
  6. Post Flight:
    • Retrieve kubeconfig and save it to jumphost

Note: Time for the above procedure is about 1.5hr, on Bare Metal Servers.

COMING SOON:

Apply a workload configuration manifests: Persistent Storage Advanced Cluster Manager Creation of the ACM Pods on the Management Cluster(Workers). Using the ACM in order to deploy OpenShift on the DU & CU servers.