OpenInfra Summit Vancouver 2023 StarlingX

Getting Started

For this hands on the machines have been pre-installed in order to save time. The image we used is located at the StarlingX Mirror Sites.

Validate access to a machine

Login to the node from IPv4 address: See StarlingX Target List for IPv4 Port numbers

ssh [email protected] -p <SSH PROXY PORT> cat /etc/build.info
Password: St8rlingX*

---

AIO-Simplex Steps

The StarlingX AIO-SX Install Guide provides details for installing an AIO SX. The steps required are listed below.

Steps :

1. Prep the node for Installation
   
2. Install StarlingX on Controller-0
   
3. Bootstrap Controller-0
   3.1. Create localhost.yml override file
   3.2. Run ansible-playbook
   
4. Configure Controller-0
   
5. Unlock Controller-0
   

For the purpose of time, steps 1 and 2 have been completed.

From the Controller, Bootstrap the system

NOTE: You must be sysadmin the password is St8rlingX*

The ansible bootstrap process checks the /home/sysadmin/loclhost.yml for the override values (more on this after we start the ansible-playbook). Make sure this file exists on your controller and then run the playbook.

export ANSIBLE_LOG_PATH=~/ansible_$(date "+%Y%m%d%H%M%S").log
time ansible-playbook /usr/share/ansible/stx-ansible/playbooks/bootstrap.yml

Bootstrap Expected results:

Below is the final output of the playbook. Note 0 failed plays

real    33m1.924s
user    14m28.441s
sys     1m43.510s
...
PLAY RECAP *********************************************************************
localhost                  : ok=465  changed=261  unreachable=0    failed=0    skipped=446  rescued=0    ignored=0   

Tuesday 13 June 2023  00:20:38 +0000 (0:00:00.457)       0:33:39.459 ********** 
=============================================================================== 
bootstrap/persist-config : Wait for service endpoints reconfiguration to complete - 381.91s
bootstrap/apply-manifest : Applying puppet bootstrap manifest --------- 374.91s
bootstrap/bringup-essential-services : Wait for 120 seconds to ensure kube-system pods are all started - 120.55s
bootstrap/bringup-essential-services : Load image from archive /opt/platform-backup/22.12/container-image1.tar.gz - 119.67s
bootstrap/persist-config : Wait for sysinv inventory ------------------- 62.54s
bootstrap/bringup-bootstrap-applications : Wait until application is in the applied state -- 58.70s
bootstrap/persist-config : Find old registry secrets in Barbican ------- 51.77s
bootstrap/bringup-essential-services : Load image from archive /opt/platform-backup/22.12/container-image2.tar.gz -- 46.83s
bootstrap/validate-config : Generate config ini file for python sysinv db population script -- 37.56s
bootstrap/bringup-bootstrap-applications : Wait until application is in the applied state -- 32.71s
bootstrap/bringup-bootstrap-applications : pause ----------------------- 30.02s
bootstrap/bringup-essential-services : Check controller-0 is in online state -- 20.44s
bootstrap/bringup-essential-services : Add loopback interface ---------- 19.91s
bootstrap/persist-config : Saving config in sysinv database ------------ 19.90s
bootstrap/bringup-bootstrap-applications : Save the current system and network config for reference in subsequent replays -- 19.66s
common/fluxcd-controllers : Get wait tasks results --------------------- 13.86s
common/bringup-kubemaster : Initializing Kubernetes master ------------- 13.61s
bootstrap/persist-config : Restart sysinv-agent and sysinv-api to pick up sysinv.conf update -- 12.20s
bootstrap/bringup-essential-services : Load image from archive /opt/platform-backup/22.12/container-image3.tar.gz -- 11.44s
bootstrap/bringup-essential-services : Pause a few seconds for docker to clean up its temp directory -- 10.02s

About the ansible override file: /home/sysadmin/localhost.yml

This file overrides default values found in /usr/share/ansible/stx-ansible/playbooks/host_vars/bootstrap/default.yml. For the purposes of this workshop, we are overriding system and network properties anything in the default.yml file can be overridden in the localhost.yml file.
By default the localhost.yml file is expected to be in the /home/sysadmin directory.

system_mode: simplex
name: "c3sxda-tc9"
description: "STX8 Standalone SX"
location: "Dallas"
contact: "[email protected]"
timezone: UTC

dns_servers:
  - "2001:4860:4860::8888"
  - "2001:4860:4860::8844"

external_oam_subnet: "2604:1380:4642:a300::100/124"
external_oam_gateway_address: "2604:1380:4642:a300::101"
external_oam_floating_address: "2604:1380:4642:a300::102"

management_subnet: "fd00:4888:2000:1090::/64"
management_start_address: "fd00:4888:2000:1090::a"
management_end_address: "fd00:4888:2000:1090::ffff"
management_multicast_subnet: "ff05::18:1:0/64"

cluster_host_subnet: "fd00:4888:2000:109a::/64"
cluster_pod_subnet: "fd00:4888:2000:109b::/64"
cluster_service_subnet: "fd00:4888:2000:109b::/112"

admin_password: St8rlingX*
ansible_become_pass: St8rlingX*

additional_local_registry_images:
  - windse/cyclictest:v1.0
  - kubernetesui/dashboard:v2.7.0
  - kubernetesui/metrics-scraper:v1.0.8

From the Controller, Configure controller-0

In this step we will configure the node for the hands-on labs. To this end, we will configure the system with Persistent storage, and isolated CPU's.

From the Controller, Configure the OAM network address

For a AIO Simplex the networking is very simple. All that is required to define is the OAM. The MGMT and Cluster host are automatically assigned to the loopback interface.

• Source the environment: source /etc/platform/openrc

• Determine the OAM device interface:

ip -6 r  | grep ^default
default via 2604:1380:4642:a300::151 dev enp1s0f0 metric 1024 onlink pref medium

Use the above output to set the OAMIF below

• Assign network interface to OAM network

OAMIF=enp1s0f0

system host-if-modify controller-0 $OAMIF -n oam0 -c platform
system interface-network-assign controller-0 oam0 oam
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org

Unlock controller

Unlocking the host will reboot the system and apply the configuration provided.

NOTE: In other scenarios more configuration can be applied before the unlock. For example configuration data networks, huge pages, persistent storage, and CPU isolated cores.

system host-unlock controller-0

Verify the System us unlocked and in a good state

NOTE: This will take a few minutes after the reboot to finish initializing

source /etc/platform/openrc
system host-list

Expected output

+----+--------------+-------------+----------------+-------------+--------------+
| id | hostname     | personality | administrative | operational | availability |
+----+--------------+-------------+----------------+-------------+--------------+
| 1  | controller-0 | controller  | unlocked       | enabled     | available    |
+----+--------------+-------------+----------------+-------------+--------------+

Verify the system is alarm free

If alarms exist the fm alarm-list command will follow up with the alarms. If no alarms are present nothing will show.

source /etc/platform/openrc
fm alarm-list

See what kubernetes pods are running

kubectl get pods -A

Expected Output

NAMESPACE      NAME                                             READY  STATUS   RESTARTS      AGE
armada         armada-api-5547f9c8d5-qd9pd                      2/2    Running  2 (6d7h ago)  6d7h
cert-manager   cm-cert-manager-6c47f6d6f5-9hcgh                 1/1    Running  1 (6d7h ago)  6d7h
cert-manager   cm-cert-manager-cainjector-6f8dc8f64d-wjktw      1/1    Running  1 (6d7h ago)  6d7h
cert-manager   cm-cert-manager-webhook-556b7d64d8-ph6s7         1/1    Running  1 (6d7h ago)  6d7h
flux-helm      helm-controller-759b895dbb-chbrb                 1/1    Running  1 (6d7h ago)  6d7h
flux-helm      source-controller-7f4bb65f88-gs7xg               1/1    Running  1 (6d7h ago)  6d7h
kube-system    calico-kube-controllers-567d594786-qsbv2         1/1    Running  1 (6d7h ago)  6d7h
kube-system    calico-node-dvmbr                                1/1    Running  1 (6d7h ago)  6d7h
kube-system    coredns-78dd5d75bd-bhncn                         1/1    Running  1 (6d7h ago)  6d7h
kube-system    ic-nginx-ingress-ingress-nginx-controller-s6jpg  1/1    Running  1 (6d7h ago)  6d7h
kube-system    kube-apiserver-controller-0                      1/1    Running  1 (6d7h ago)  6d7h
kube-system    kube-controller-manager-controller-0             1/1    Running  1 (6d7h ago)  6d7h
kube-system    kube-multus-ds-amd64-d2t4j                       1/1    Running  1 (6d7h ago)  6d7h
kube-system    kube-proxy-tbklg                                 1/1    Running  1 (6d7h ago)  6d7h
kube-system    kube-scheduler-controller-0                      1/1    Running  1 (6d7h ago)  6d7h
kube-system    kube-sriov-cni-ds-amd64-fzzqd                    1/1    Running  1 (6d7h ago)  6d7h

Next Exercises

---

• StarlingX Management overview
• Cyclic Test
• StarlingX Hello World
• Kubernetes Dashboard
• Persistent Storage
• Install Matrix Server

---

Exercises for Home

---

• Running StarlingX on libvirt
• Running StarlingX on virtualbox

---

Hands On Setup

---

1. Equinix Metal starlingX hands-on Workspace setup
2. Access StarlingX Targets through Jumphost
3. To Install on equinix-metal see StarlingX on Demand Metal