Use Ceph monitoring (not RGW) network for Dashboard

Use the existing `cifmw_cephadm_monitoring_network` parameter to configure the Ceph dashboard instead of the `cifmw_cephadm_rgw_network` parameter. This was not detected earlier because the Ceph playbook set them both to the Ceph public_network, but we recently added a feature to run RGW on a separate network [1]. Also, add retries/delay to the dashboard HTTP check. [1] openstack-k8s-operators#2301 Jira: https://issues.redhat.com/browse/OSPCIX-506 Signed-off-by: John Fulton <[email protected]>
fultonj · Sep 26, 2024 · 2a06d3f · 2a06d3f
1 parent 2520f74
commit 2a06d3f
Show file tree

Hide file tree

Showing 4 changed files with 13 additions and 4 deletions.
diff --git a/roles/cifmw_cephadm/defaults/main.yml b/roles/cifmw_cephadm/defaults/main.yml
@@ -138,3 +138,5 @@ cifmw_cephadm_update_log_commands:
   # Get last cephadm logs in case of failure
   - type: "mod_cephadm"
     cmd: "log last cephadm"
+cifmw_cephadm_wait_for_dashboard_retries: 10
+cifmw_cephadm_wait_for_dashboard_delay: 20
diff --git a/roles/cifmw_cephadm/tasks/dashboard/configure_dashboard_backends.yml b/roles/cifmw_cephadm/tasks/dashboard/configure_dashboard_backends.yml
@@ -34,5 +34,5 @@
   ansible.builtin.command: |
     {{ cifmw_cephadm_ceph_cli }} config set \
     mgr mgr/dashboard/{{ current_mgr }}/server_addr \
-    {{ hostvars[item][all_addresses] | ansible.utils.ipaddr(cifmw_cephadm_rgw_network) | first }}
+    {{ hostvars[item][all_addresses] | ansible.utils.ipaddr(cifmw_cephadm_monitoring_network) | first }}
   become: true
diff --git a/roles/cifmw_cephadm/tasks/dashboard/validation.yml b/roles/cifmw_cephadm/tasks/dashboard/validation.yml
@@ -4,6 +4,10 @@
     dest: "/tmp/dash_response"
   register: dashboard_response
   failed_when: dashboard_response.failed == true
+  until: dashboard_response.failed == false
+  retries: "{{ cifmw_cephadm_wait_for_dashboard_retries }}"
+  delay: "{{ cifmw_cephadm_wait_for_dashboard_delay }}"
+
 
 - name: Check http response code from dashboard service with login
   ansible.builtin.get_url:
@@ -13,3 +17,6 @@
     password: admin
   register: dashboard_resp
   failed_when: dashboard_resp.status_code != 200
+  until: dashboard_resp.status_code == 200
+  retries: "{{ cifmw_cephadm_wait_for_dashboard_retries }}"
+  delay: "{{ cifmw_cephadm_wait_for_dashboard_delay }}"
diff --git a/roles/cifmw_cephadm/templates/ceph_monitoring_stack.yml.j2 b/roles/cifmw_cephadm/templates/ceph_monitoring_stack.yml.j2
@@ -11,7 +11,7 @@ placement:
   - {{ _hosts[0] }}
   count: 1
 networks:
-- {{ cifmw_cephadm_rgw_network }}
+- {{ cifmw_cephadm_monitoring_network }}
 ---
 service_type: prometheus
 service_id: prometheus
@@ -21,7 +21,7 @@ placement:
   - {{ _hosts[0] }}
   count: 1
 networks:
-- {{ cifmw_cephadm_rgw_network }}
+- {{ cifmw_cephadm_monitoring_network }}
 ---
 service_type: alertmanager
 service_id: alertmanager
@@ -31,4 +31,4 @@ placement:
   - {{ _hosts[0] }}
   count: 1
 networks:
-- {{ cifmw_cephadm_rgw_network }}
+- {{ cifmw_cephadm_monitoring_network }}