pvc pending

[infordb]

Hi
I am testing portworx on GKE
I following your example of cassandra (https://docs.portworx.com/scheduler/kubernetes/cassandra-k8s.html)

The first pod(cassandra-0) is assigned a volume well, but second pod(cassandra-1) is not
pvc is still pending
What should I check more ?

infordb_park@speedy-area-195923:~$ kubectl get pvc
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
cassandra-data-cassandra-0 Bound pvc-b0d2eb75-1f68-11e8-b061-42010a9200c5 1Gi RWO px-storageclass 22h
cassandra-data-cassandra-1 Pending px-storageclass 22h

infordb_park@speedy-area-195923:~$ kubectl describe pvc cassandra-data-cassandra-1
Name: cassandra-data-cassandra-1
Namespace: default
StorageClass: px-storageclass
Status: Pending
Volume:
Labels: app=cassandra
Annotations: volume.beta.kubernetes.io/storage-class=px-storageclass
volume.beta.kubernetes.io/storage-provisioner=kubernetes.io/portworx-volume
Capacity:
Access Modes:
Events:
Type Reason Age From Message

---

Warning ProvisioningFailed 2m (x5382 over 22h) persistentvolume-controller Failed to provision volume with StorageClass "px-storageclass": Could not find enough online nodes with sufficient storage at requested Cos (HIGH) level. 2 node(s)
were excluded to enforce group provisioning.
Warning ProvisioningFailed 2m (x5382 over 22h) persistentvolume-controller Failed to provision volume with StorageClass "px-storageclass": Get http://10.55.252.27:9001/v1/osd-volumes/versions: dial tcp 10.55.252.27:9001: i/o timeout
infordb_park@speedy-area-195923:~$

root@gke-cluster-1-default-pool-ea79223e-6srv:~# /opt/pwx/bin/pxctl status
Status: PX is operational
License: Trial (expires in 30 days)
Node ID: gke-cluster-1-default-pool-ea79223e-6srv
IP: 10.146.0.5
Local Storage Pool: 1 pool
POOL IO_PRIORITY RAID_LEVEL USABLE USED STATUS ZONE REGION
0 LOW raid0 10 GiB 2.1 GiB Online default default
Local Storage Devices: 1 device
Device Path Media Type Size Last-Scan
0:1 /dev/sdb STORAGE_MEDIUM_MAGNETIC 10 GiB 04 Mar 18 04:30 UTC
total - 10 GiB
Cluster Summary
Cluster ID: mycluster
Cluster UUID: df97b45c-a095-4568-bd55-a120badaf66a
Nodes: 3 node(s) with storage (3 online)
IP ID StorageNode Used Capacity Status
10.146.0.3 gke-cluster-1-default-pool-ea79223e-n81h Yes 2.1 GiB 10 GiB Online
10.146.0.2 gke-cluster-1-default-pool-ea79223e-gqkr Yes 2.0 GiB 10 GiB Online
10.146.0.5 gke-cluster-1-default-pool-ea79223e-6srv Yes 2.1 GiB 10 GiB Online (This node)
Global Storage Pool
Total Used : 6.2 GiB
Total Capacity : 30 GiB

[infordb]

If I remove the fg: "true" option, it will work well.
what is this option ? Is this option essential?

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
name: px-storageclass
provisioner: kubernetes.io/portworx-volume
parameters:
repl: "2"
priority_io: "high"
group: "cassandra_vg"
fg: "true"

[sangleganesh]

When you specify "group" option, it means try to allocate nodes for the volume data where no other volumes exist created with the same "group" option. If it cannot find enough nodes, it retries without group option.
The "fg" option is enforce group - i.e. fail volume create if not enough nodes exist where no volumes with given group exist.

[venkatpx]

@infordb were you able to proceed with the above?