snc
snc copied to clipboard
crc-org
Setup kubevirt-hostpath-provisioner
this sets up dynamic provisioner for pvs backed by hosts's filesystem
the CSI driver is part of the kubevirt project and the manifests used are available in their repo at: https://github.com/kubevirt/hostpath-provisioner/tree/main/deploy/csi
this also modifies the pvc manifest for the internal registry to set the StorageClass that uses the CSI based provisioner
when creating new pvc one should use the storageClassName "crc-csi-hostpath-provisioner"
Fixes #444
about resource consumption in the cluster, with just the plain dynamic provisioner the cluster doesn't seem to consume more resource, while deploying the operator it takes around ~750mb more memory
# CRC with the hostpath-provisioner (deployed as an old provisioner)
Capacity:
cpu: 5
ephemeral-storage: 31970284Ki
hugepages-1Gi: 0
hugepages-2Mi: 0
memory: 13999196Ki
pods: 250
Allocatable:
cpu: 4800m
ephemeral-storage: 29463813686
hugepages-1Gi: 0
hugepages-2Mi: 0
memory: 13538396Ki
pods: 250
Allocated resources:
(Total limits may be over 100 percent, i.e., overcommitted.)
Resource Requests Limits
-------- -------- ------
cpu 2250m (46%) 0 (0%)
memory 7392Mi (55%) 0 (0%)
# CRC with the hostpath-provisioner-operator (deployed as CSI driver managed by the operator)
Capacity:
cpu: 5
ephemeral-storage: 31970284Ki
hugepages-1Gi: 0
hugepages-2Mi: 0
memory: 13999204Ki
pods: 250
Allocatable:
cpu: 4800m
ephemeral-storage: 29463813686
hugepages-1Gi: 0
hugepages-2Mi: 0
memory: 13538404Ki
pods: 250
Allocated resources:
(Total limits may be over 100 percent, i.e., overcommitted.)
Resource Requests Limits
-------- -------- ------
cpu 2300m (47%) 0 (0%)
memory 8142Mi (61%) 0 (0%)
after much hit and trial, CSIDriver deployment is working now, finally :slightly_smiling_face: the provided manifests don't work ootb, resource consumption is minimal compared to the operator
[APPROVALNOTIFIER] This PR is APPROVED
This pull-request has been approved by: anjannath
The full list of commands accepted by this bot can be found here.
The pull request process is described here
- ~~OWNERS~~ [anjannath]
Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment
![openshift-ci[bot] avatar](https://avatars.githubusercontent.com/in/91280?v=4 "Published by a pub.dev verified publisher"){kind=small}
Somehow the CI failed because it detect that hostpath-provisioner namespace already present. I think you should add the namespace resource to one of the yaml or create a separate one and use oc apply to create it so that if you run same apply multiple time it doesn't change the resource and also not exist with non-zero.
$ cat host.yaml
apiVersion: v1
kind: Namespace
metadata:
name: hostpath-provisioner
$ oc apply -f host.yaml
namespace/hostpath-provisioner unchanged
downstream snc job passed with 4.11.10 not sure what is the reason for consistent failure with the openshift CI job :/
the image registry pvc requests volume with capacity: 100Gi the CSI driver needs actual free disk space for this and since the VM created is only 30GB it fails to create the pv
as this was working previously when manually creating the pv with the right size, we might have to do that for the registry operator and the csi driver will be used for user workloads
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal WaitForPodScheduled 81s (x2 over 83s) persistentvolume-controller waiting for pod image-registry-84489d56b9-5wzhh to be scheduled
Normal Provisioning 15s (x7 over 78s) kubevirt.io.hostpath-provisioner_csi-hostpathplugin-2lknn_f802a19c-3b61-4624-ae77-348077c21b0b External provisioner is provisioning volume for claim "openshift-image-registry/crc-image-registry-storage"
Warning ProvisioningFailed 15s (x7 over 78s) kubevirt.io.hostpath-provisioner_csi-hostpathplugin-2lknn_f802a19c-3b61-4624-ae77-348077c21b0b failed to provision volume with StorageClass "crc-csi-hostpath-provisioner": created volume capacity 32212254720 less than requested capacity 107374182400
Normal ExternalProvisioning 6s (x7 over 78s) persistentvolume-controller waiting for a volume to be created, either by external provisioner "kubevirt.io.hostpath-provisioner" or manually created by system administrator
@praveenkumar: only crc-org org members may request cherry-picks. You can still do the cherry-pick manually.
In response to this:
/cherry-pick release-4.12
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.
@praveenkumar: new pull request created: #594
In response to this:
/cherry-pick release-4.12
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.