ibm-storage-odf-console provides IBM storage specific console page, which will be loaded by ODF console when end users access IBM storage. It's specially designed for displaying IBM specific storage attributes to customer. Current scope includes IBM flashsystem only.
Apache License 2.0
3
stars
7
forks
source link
Spectrum scale core and gui pods stuck in init. Missing secret. #39
All of the scale core pods and gui pods are stuck in init. There are no obvious errors in the gui pod. The scale core pod mention a missing secret. I assume it is the operator that creates the secret.
oc describe says that a secret doesn't exist.
But our instructions didn't say to create a secret. I think the operator is doing that b/c the name of the secret that it is looking for is not something I would have chosen: scale-core-token-87bkd
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning FailedMount 14m (x4655 over 6d8h) kubelet (combined from similar events): Unable to attach or mount volumes: unmounted volumes=[scale-core-token-87bkd], unattached volumes=[asm-generic modules gpfs0 scale-core-token-87bkd gpfssystemlog core-config scale-pmsensors-volume asm linux-include gpfslog gui-config kernels ssh-keys mmfs overlay]: timed out waiting for the condition
Warning FailedMount 3m48s (x3890 over 6d8h) kubelet MountVolume.SetUp failed for volume "scale-core-token-87bkd" : secret "scale-core-token-87bkd" not found
[root@cluster2-inf ~]#
On a side note, should CSI start before Spectrum Scale? Maybe that is OK but not sure. In previous projects, CNSA (cloud native Spectrum Scale), CSI had a dependency on Spectrum scale . Anyway that is not what I am reporting here.
What is at issue here is that the Spectrum Scale pods are stuck in Init.
All of the scale core pods and gui pods are stuck in init. There are no obvious errors in the gui pod. The scale core pod mention a missing secret. I assume it is the operator that creates the secret.
oc describe says that a secret doesn't exist. But our instructions didn't say to create a secret. I think the operator is doing that b/c the name of the secret that it is looking for is not something I would have chosen: scale-core-token-87bkd
These are my secrets.
CSI pods are all running with no problems.
On a side note, should CSI start before Spectrum Scale? Maybe that is OK but not sure. In previous projects, CNSA (cloud native Spectrum Scale), CSI had a dependency on Spectrum scale . Anyway that is not what I am reporting here.
What is at issue here is that the Spectrum Scale pods are stuck in Init.