Merge pull request #1313 from Zarquan/20240219-zrq-jade-debug

20240219 zrq jade debug
wfau · Feb 29, 2024 · 10a3c5a · 10a3c5a
2 parents c91e01a + f142312
commit 10a3c5a
Show file tree

Hide file tree

Showing 13 changed files with 2,846 additions and 8 deletions.
diff --git a/bin/kube-client b/bin/kube-client
@@ -65,6 +65,7 @@
     containerfull=ghcr.io/wfau/atolmis/${containername:?}
 
     kubectlproxy=127.0.0.1::8001
+    monitorproxy=127.0.0.1::3001
 
     podman run \
         --rm \
@@ -73,6 +74,7 @@
         --name     "${clientname:?}" \
         --hostname "${clientname:?}" \
         --publish  "${kubectlproxy:?}" \
+        --publish  "${monitorproxy:?}" \
         --env "cloudname=${cloudname:?}" \
         --env "cloudsite=${cloudsite:?}" \
         --env "SSH_AUTH_SOCK=/mnt/ssh_auth_sock" \

diff --git a/deployments/cluster-api/ansible/00-create-all.yml b/deployments/cluster-api/ansible/00-create-all.yml
@@ -29,7 +29,7 @@
 - import_playbook: 22-install-capi-provider.yml
 - import_playbook: 23-install-capi-helm-charts.yml
 - import_playbook: 25-create-work-cluster.yml
-- import_playbook: 26-secure-work-cluster.yml
+# import_playbook: 26-secure-work-cluster.yml
 
 - import_playbook: 30-install-aglais.yml
 

diff --git a/deployments/cluster-api/ansible/21-create-kind-cluster.yml b/deployments/cluster-api/ansible/21-create-kind-cluster.yml
@@ -46,7 +46,7 @@
         dest: "{{ aglais.kubernetes.cluster.kind.conf }}"
         flat: yes
 
-- name: "Set local file permissions"
+- name: "Update localhost"
   gather_facts: false
   hosts: localhost
   vars_files:
@@ -61,3 +61,11 @@
         path: "{{ aglais.kubernetes.cluster.kind.conf }}"
         mode: "u=rw,g=,o="
 
+    - name: "Update timestamp in [{{ agstatusfile }}]"
+      yedit:
+        src: "{{ agstatuspath }}"
+        key: aglais.kubernetes.cluster.kind.debug
+        value:
+          created: "{{ now('%Y-%m-%dT%H:%M:%S%:z') }}"
+
+
diff --git a/deployments/cluster-api/ansible/25-create-work-cluster.yml b/deployments/cluster-api/ansible/25-create-work-cluster.yml
@@ -77,7 +77,7 @@
         dest: "{{ aglais.kubernetes.cluster.work.conf }}"
         flat: yes
 
-- name: "Set local file permissions"
+- name: "Update localhost"
   gather_facts: false
   hosts: localhost
   vars_files:
@@ -92,4 +92,11 @@
         path: "{{ aglais.kubernetes.cluster.work.conf }}"
         mode: "u=rw,g=,o="
 
+    - name: "Update timestamp in [{{ agstatusfile }}]"
+      yedit:
+        src: "{{ agstatuspath }}"
+        key: aglais.kubernetes.cluster.work.debug
+        value:
+          created: "{{ now('%Y-%m-%dT%H:%M:%S%:z') }}"
+
 
diff --git a/deployments/cluster-api/ansible/templates/clusterapi-config.j2 b/deployments/cluster-api/ansible/templates/clusterapi-config.j2
@@ -93,8 +93,7 @@ addons:
 
   # Settings for the CNI addon
   cni:
-
-    # Indicates if a CNI should be deployed
+    # Indicates if a CNI should be deployed (default true)
     enabled: true
 
     # The CNI to deploy - supported values are calico or cilium
@@ -108,8 +107,53 @@ addons:
         name: tigera-operator
         version: v3.26.0
 
-  # Include the Kubernetes dashboard
+  # Settings for the OpenStack integrations
+  openstack:
+    # Indicates if the OpenStack integrations should be enabled (default false)
+    enabled: false
+
+  # Settings for the metrics server
+  # https://github.com/kubernetes-sigs/metrics-server#helm-chart
+  metricsServer:
+    # Indicates if the metrics server should be deployed (default true)
+    enabled: true
+
+  # Settings for the Kubernetes dashboard
+  # https://github.com/kubernetes/dashboard/tree/master/charts/helm-chart/kubernetes-dashboard
   kubernetesDashboard:
+    # Indicates if the Kubernetes dashboard should be enabled (default false)
+    enabled: true
+
+  # Settings for ingress controllers
+  ingress:
+    # Indicates if ingress controllers should be enabled (default false)
+    enabled: false
+
+  # Settings for cluster monitoring
+  monitoring:
+    # Indicates if the cluster monitoring should be enabled (default false)
+    enabled: true
+
+  # Settings for node feature discovery
+  # https://github.com/kubernetes-sigs/node-feature-discovery/tree/master/deployment/helm/node-feature-discovery
+  nodeFeatureDiscovery:
+    # Indicates if node feature discovery should be enabled (default true)
+    enabled: true
+
+  # Settings for the NVIDIA GPU operator
+  nvidiaGPUOperator:
+    # Indicates if the NVIDIA GPU operator should be enabled (default true)
+    # Note that because it uses node feature discovery to run only on nodes
+    # with an NVIDIA GPU available, the overhead of enabling this on clusters
+    # that do not need it now but may need it in the future is low
+    enabled: true
+
+  # Settings for the Mellanox network operator
+  mellanoxNetworkOperator:
+    # Indicates if the network operator should be enabled (default true)
+    # Note that because it uses node feature discovery to run only on nodes
+    # with a Mellanox NIC available, the overhead of enabling this on clusters
+    # that do not need it now but may need it in the future is low
     enabled: true
 
 
diff --git a/deployments/cluster-api/ansible/templates/init-status.j2 b/deployments/cluster-api/ansible/templates/init-status.j2
@@ -26,6 +26,8 @@ aglais:
     type: cluster-api
     name: {{ deployname }}
     date: {{ deploydate }}
+    debug:
+      started: "{{ now('%Y-%m-%dT%H:%M:%S%:z') }}"
   openstack:
     cloud:
       name: {{ cloudname }}

diff --git a/notes/zrq/20240219-03-jade-reconnect.txt b/notes/zrq/20240219-03-jade-reconnect.txt
@@ -61,12 +61,10 @@
         '
         date
         hostname
-        ls -al /opt/aglais/aglais-status.yml
         '
 
     >   Mon Feb 19 05:58:30 PM UTC 2024
     >   somerville-jade-20240219-bootstrap-node.novalocal
-    >   -rw-r--r--. 1 root root 1970 Feb 19 17:02 /opt/aglais/aglais-status.yml
 
 
 # -----------------------------------------------------
@@ -112,3 +110,42 @@
     # Client configured :-D
     #
 
+
+# -----------------------------------------------------
+# Fetch the kubectl configuration files.
+#[root@ansibler]
+
+    yq '.aglais.kubernetes.cluster.kind.conf' \
+        '/opt/aglais/aglais-status.yml'
+
+    >   /opt/aglais/somerville-jade-20240221-kind.yml
+
+
+    yq '.aglais.kubernetes.cluster.work.conf' \
+        '/opt/aglais/aglais-status.yml'
+
+    >   /opt/aglais/somerville-jade-20240221-work.yml
+
+
+    kindclusterconf=$(
+        yq '.aglais.kubernetes.cluster.kind.conf' \
+            '/opt/aglais/aglais-status.yml'
+        )
+
+    workclusterconf=$(
+        yq '.aglais.kubernetes.cluster.work.conf' \
+            '/opt/aglais/aglais-status.yml'
+        )
+
+    scp "root@bootstrap:${kindclusterconf}" \
+        "${kindclusterconf}"
+
+    scp "root@bootstrap:${workclusterconf}" \
+        "${workclusterconf}"
+
+    #
+    # This doesn't work because the kubectl API is blocked by Somerville firewall.
+    # Only ssh is allowed to public IP addresses ?
+    #
+
+