fix: rook-ceph metrics error

ahgraber · Jan 29, 2024 · 51a2821 · 51a2821
1 parent 090a6a1
commit 51a2821
Show file tree

Hide file tree

Showing 4 changed files with 22 additions and 2 deletions.
diff --git a/kubernetes/apps/monitoring/kube-prometheus-stack/app/kube-state-metrics.yaml b/kubernetes/apps/monitoring/kube-prometheus-stack/app/kube-state-metrics.yaml
@@ -12,7 +12,7 @@ kube-state-metrics:
       relabelings:
         - action: replace
           sourceLabels: ["__meta_kubernetes_pod_node_name"]
-          regex: ^(.*)$
+          regex: (.*)
           replacement: $1
           targetLabel: kubernetes_node
   rbac:

diff --git a/kubernetes/apps/rook-ceph/README.md b/kubernetes/apps/rook-ceph/README.md
@@ -69,6 +69,14 @@ Run `task ceph:teardown`
 
 ## Troubleshooting
 
+### Ceph reports no orchestrator configured
+
+> run the following commands against the ceph-toolbox pod
+
+```sh
+ceph mgr module enable rook && ceph orch set backend rook && ceph orch status
+```
+
 ### Dashboard not accessible thru ingress
 
 > run the following commands against the ceph-toolbox pod

diff --git a/kubernetes/apps/rook-ceph/rook-ceph/cluster/helmrelease.yaml b/kubernetes/apps/rook-ceph/rook-ceph/cluster/helmrelease.yaml
@@ -71,7 +71,7 @@ spec:
       [global]
       bdev_enable_discard = true
       bdev_async_discard = true
-
+      osd_class_update_on_start = false
     cephClusterSpec:
       crashCollector:
         disable: false
@@ -86,6 +86,15 @@ spec:
         ssl: false
       network:
         provider: host
+      placement:
+        mgr: &placement
+          nodeAffinity:
+            requiredDuringSchedulingIgnoredDuringExecution:
+              nodeSelectorTerms:
+                - matchExpressions:
+                    - key: node-role.kubernetes.io/control-plane
+                      operator: Exists
+        mon: *placement
       storage:
         useAllNodes: false
         useAllDevices: false

diff --git a/kubernetes/apps/rook-ceph/rook-ceph/operator/helmrelease.yaml b/kubernetes/apps/rook-ceph/rook-ceph/operator/helmrelease.yaml
@@ -48,6 +48,9 @@ spec:
         operator: Exists
 
     csi:
+      enableLiveness: true
+      serviceMonitor:
+        enabled: true
       # enableRbdDriver: true
       # enableCephfsDriver: true
       # enableGrpcMetrics: false