Update pv workflow example

kubernetes · Jan 30, 2017 · 2728af6 · 2728af6
1 parent 71f2016
commit 2728af6
Showing 1 changed file with 23 additions and 23 deletions.
diff --git a/contributors/design-proposals/local-storage-overview.md b/contributors/design-proposals/local-storage-overview.md
@@ -184,12 +184,12 @@ spec:
 ```yaml
 $ kubectl get pv
 NAME       CAPACITY ACCESSMODES RECLAIMPOLICY STATUS    CLAIM … NODE  
-local-pv-1 375Gi    RWO         Delete        Available         gke-mycluster-1
-local-pv-2 375Gi    RWO         Delete        Available         gke-mycluster-1
-local-pv-1 375Gi    RWO         Delete        Available         gke-mycluster-2
-local-pv-2 375Gi    RWO         Delete        Available         gke-mycluster-2
-local-pv-1 375Gi    RWO         Delete        Available         gke-mycluster-3
-local-pv-2 375Gi    RWO         Delete        Available         gke-mycluster-3
+local-pv-1 100Gi    RWO         Delete        Available         node-1
+local-pv-2 10Gi     RWO         Delete        Available         node-1
+local-pv-1 100Gi    RWO         Delete        Available         node-2
+local-pv-2 10Gi     RWO         Delete        Available         node-2
+local-pv-1 100Gi    RWO         Delete        Available         node-3
+local-pv-2 10Gi     RWO         Delete        Available         node-3
 ```
 3. The addon will monitor the health of secondary partitions and taint PVs whenever the backing local storage devices becomes unhealthy.
 4. Alice creates a StatefulSet that uses local PVCs
@@ -222,19 +222,19 @@ spec:
   - metadata:
       name: www
       labels:
-        storage.kubernetes.io/medium: local-ssd
-        storage.kubernetes.io/volume-type: local
+        storage.kubernetes.io/medium: ssd
     spec:
+      volume-type: local
       accessModes: [ "ReadWriteOnce" ]
       resources:
         requests:
-          storage: 1Gi
+          storage: 100Gi
   - metadata:
       name: log
       labels:
         storage.kubernetes.io/medium: hdd
-        storage.kubernetes.io/volume-type: local
     spec:
+      volume-type: local
       accessModes: [ "ReadWriteOnce" ]
       resources:
         requests:
@@ -244,22 +244,22 @@ spec:
 ```yaml
 $ kubectl get pvc
 NAME            STATUS VOLUME     CAPACITY ACCESSMODES … NODE  
-www-local-pvc-1 Bound  local-pv-1 375Gi    RWO           gke-mycluster-1
-www-local-pvc-2 Bound  local-pv-1 375Gi    RWO           gke-mycluster-2
-www-local-pvc-3 Bound  local-pv-1 375Gi    RWO           gke-mycluster-3
-log-local-pvc-1 Bound  local-pv-1 375Gi    RWO           gke-mycluster-1
-log-local-pvc-2 Bound  local-pv-1 375Gi    RWO           gke-mycluster-2
-log-local-pvc-3 Bound  local-pv-1 375Gi    RWO           gke-mycluster-3
+www-local-pvc-1 Bound  local-pv-1 375Gi    RWO           node-1
+www-local-pvc-2 Bound  local-pv-1 375Gi    RWO           node-2
+www-local-pvc-3 Bound  local-pv-1 375Gi    RWO           node-3
+log-local-pvc-1 Bound  local-pv-2 375Gi    RWO           node-1
+log-local-pvc-2 Bound  local-pv-2 375Gi    RWO           node-2
+log-local-pvc-3 Bound  local-pv-2 375Gi    RWO           node-3
 ```
 ```yaml
 $ kubectl get pv
 NAME       CAPACITY … STATUS    CLAIM           NODE  
-local-pv-1 375Gi      Bound     www-local-pvc-1 gke-mycluster-1
-local-pv-2 375Gi      Bound     log-local-pvc-1 gke-mycluster-1
-local-pv-1 375Gi      Bound     www-local-pvc-2 gke-mycluster-2
-local-pv-2 375Gi      Bound     log-local-pvc-2 gke-mycluster-2
-local-pv-1 375Gi      Bound     www-local-pvc-3 gke-mycluster-3
-local-pv-2 375Gi      Bound     log-local-pvc-3 gke-mycluster-3
+local-pv-1 100Gi      Bound     www-local-pvc-1 node-1
+local-pv-2 10Gi       Bound     log-local-pvc-1 node-1
+local-pv-1 100Gi      Bound     www-local-pvc-2 node-2
+local-pv-2 10Gi       Bound     log-local-pvc-2 node-2
+local-pv-1 100Gi      Bound     www-local-pvc-3 node-3
+local-pv-2 10Gi       Bound     log-local-pvc-3 node-3
 ```
 6. If a pod dies and is replaced by a new one that reuses existing PVCs, the pod will be placed on the same node where the corresponding PVs exist. Stateful Pods are expected to have a high enough priority which will result in such pods preempting other low priority pods if necessary to run on a specific node.
 7. If a new pod fails to get scheduled while attempting to reuse an old PVC, the StatefulSet controller is expected to give up on the old PVC (delete & recycle) and instead create a new PVC based on some policy. This is to guarantee scheduling of stateful pods.
@@ -290,7 +290,7 @@ metadata:
   annotations:
     storage.kubernetes.io/node: k8s-node
   labels:
-    storage.kubernetes.io/medium: hdd
+    storage.kubernetes.io/medium: ssd
 spec:
   volume-type: local
   storage-type: block