feat: force pods with volumes to be scheduled on Cloud servers (#743)

Due to a bug in the scheduler a node with no driver instance might be picked and the volume is stuck in pending as the "no capacity - > reschedule" recovery is never triggered [[0]](kubernetes/kubernetes#122109), [[1]](kubernetes-csi/external-provisioner#544). - See #400 --------- Co-authored-by: lukasmetzner <[email protected]> Co-authored-by: Julian Tölle <[email protected]>
hetznercloud · Nov 11, 2024 · cd5c717 · cd5c717
1 parent 7211dd8
commit cd5c717
Show file tree

Hide file tree

Showing 6 changed files with 45 additions and 2 deletions.
diff --git a/chart/.snapshots/default.yaml b/chart/.snapshots/default.yaml
@@ -142,6 +142,10 @@ spec:
                 operator: NotIn
                 values:
                 - "true"
+              - key: instance.hetzner.cloud/provided-by
+                operator: NotIn
+                values:
+                - robot
       tolerations:
         - effect: NoExecute
           operator: Exists

diff --git a/chart/.snapshots/full.values.yaml b/chart/.snapshots/full.values.yaml
@@ -370,3 +370,5 @@ storageClasses:
   - name: foobar
     defaultStorageClass: false
     reclaimPolicy: Keep
+    allowedTopologyCloudServer: false
+
diff --git a/chart/.snapshots/full.yaml b/chart/.snapshots/full.yaml
@@ -245,6 +245,10 @@ spec:
                 operator: NotIn
                 values:
                 - "true"
+              - key: instance.hetzner.cloud/provided-by
+                operator: NotIn
+                values:
+                - robot
       nodeSelector:
         foo: bar
       tolerations:

diff --git a/chart/values.yaml b/chart/values.yaml
@@ -556,6 +556,10 @@ node:
                 operator: NotIn
                 values:
                   - "true"
+              - key: "instance.hetzner.cloud/provided-by"
+                operator: NotIn
+                values:
+                  - "robot"
 
   ## @param node.nodeSelector Node labels for node pods assignment
   ## ref: https://kubernetes.io/docs/user-guide/node-selection/

diff --git a/deploy/kubernetes/hcloud-csi.yml b/deploy/kubernetes/hcloud-csi.yml
diff --git a/docs/kubernetes/README.md b/docs/kubernetes/README.md
@@ -106,7 +106,7 @@ metadata:
 stringData:
  encryption-passphrase: foobar
 
---- 
+---
 
 apiVersion: storage.k8s.io/v1
 kind: StorageClass
@@ -209,8 +209,33 @@ $ kubectl apply -f https://raw.githubusercontent.com/hetznercloud/csi-driver/v2.
 
 ## Integration with Root Servers
 
-Root servers can be part of the cluster, but the CSI plugin doesn't work there. Taint the root server as follows to skip that node for the DaemonSet.
+Root servers can be part of the cluster, but the CSI plugin doesn't work there.
 
+Labels are needed to indicate whether a node is a cloud VM or a dedicated server from Robot. If you are using the `hcloud-cloud-controller-manager` version 1.21.0 or later, these labels are added automatically. Otherwise, you will need to label the nodes manually.
+
+### Adding labels manually
+
+**Cloud Servers**
+```bash
+kubectl label nodes <node name> instance.hetzner.cloud/provided-by=cloud
+```
+
+**Root Servers**
+```bash
+kubectl label nodes <node name> instance.hetzner.cloud/provided-by=robot
+```
+
+
+### DEPRECATED: Old Label
+
+We prefer that you use our [new label](#new-label). The label `instance.hetzner.cloud/is-robot-server` will be deprecated in future releases.
+
+**Cloud Servers**
+```bash
+kubectl label nodes <node name> instance.hetzner.cloud/is-root-server=false
+```
+
+**Root Servers**
 ```bash
 kubectl label nodes <node name> instance.hetzner.cloud/is-root-server=true
 ```