Closed triceam closed 2 years ago
is there any reason why the default storage (block) or maybe ibmc-vpc-block-metro-10iops-tier
since wait for first customer is specified in https://github.com/turbonomic/t8c-install/wiki/Storage-Class-Requirements ?
I was able to recover the deployment by changing the storage class used by Turbonomic to the cluster default: ibmc-vpc-block-10iops-tier
Before changing the automation, I need to understand why we created that storage class in the first place... and I do not know why. @tcskill @mjperrins @seansund Any idea why we create the ibmc-vpc-block-mzr
storage class for IBM Cloud deployments? This is what is causing turbo to fail.
@triceam yes. That is needed on ibm cloud only when you have a vpc deployment or multi region deployment because by default Ibm cloud creates block storage with bind immediately and need it to be wait for first consumer (note only for the use case described above) when on classic cluster not needed, also when on single zone not needed
standard metro classes will work for this use case, no need for a custom storageclass
OK, well, it doesn't work anymore. The <sizeRange> is an invalid parameter, Action: Please provide valid parameters}
error is b/c turbo is requesting 1Gi volumes, and the minimum size is 10Gi. I will update our automation to use the metro class.
Yes I see new metro has wait for first consumer so should work fine
Removed 202 layer (no custom storage class for IBM Cloud, switched to use ibmc-vpc-block-metro-10iops-tier
storage class on IBM Cloud.
Fixed in:
The turbnomic automation fails to deploy on both OCP 4.8 and 4.10 on IBM Cloud.
Automation: https://github.com/IBM/automation-turbonomic
The deployment never becomes healthy b/c all of the PVCs (which use the storage class created by the 202 layer) never become healthy.
The error events coming from the PVCs are similar to:
The storage class has a volume size 10-2000Gi. The volumes have a size of 1Gi, so I suspect that is the issue.