Iussue- During customization, the VMConfig plug-in tries to fit the virtual machines to the correct NUMA nodes of the ESXi server based on its SR-IOV, Passthrough, and Pinning specifications.
Workaround-
To disable the virtual machine placement operation by the plug-in and enable the virtual machine placement by vSphere DRS, configure the following settings on the node pool before instantiating the network function.
The following is the documented workaround:
Connect to TKG Management cluster using SSH.
Execute the following command on the selected nodepool kubectl edit vmconfigset -n test pn1
and add preferHosts: [] Example before change CCLI@tph-ci-vmw-caas-mgmt1>>kubectl get vmconfigset -n test test-np1 -o yaml
apiVersion: acm.vmware.com/v1alpha1
kind: VMConfigSet
metadata:
annotations:
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"acm.vmware.com/v1alpha1","kind":"VMConfigSet","metadata":{"annotations":{},"name":"test-np1","namespace":"test"},"spec":{"nodeLabels":{"telco.vmware.com/nodepool":"test-np1","type":"h"},"vmConfig":{"extraConfig":{"sched.cpu.latencySensitivity":"high"},"preferHosts":null,"revision":"1643191324469087192"}}}
creationTimestamp: "2022-01-24T14:20:12Z"
generation: 3
managedFields:
- apiVersion: acm.vmware.com/v1alpha1
fieldsType: FieldsV1
fieldsV1:
f:metadata:
f:annotations:
.: {}
f:kubectl.kubernetes.io/last-applied-configuration: {}
f:spec:
.: {}
f:nodeLabels:
.: {}
f:telco.vmware.com/nodepool: {}
f:type: {}
f:vmConfig:
.: {}
f:extraConfig:
.: {}
f:sched.cpu.latencySensitivity: {}
f:revision: {}
manager: kubectl-client-side-apply
operation: Update
time: "2022-01-24T15:41:32Z"
name: test-np1
namespace: test
resourceVersion: "82684707"
uid: d821db3b-e062-48d9-9cb5-e61195ec63d6
spec:
nodeLabels:
telco.vmware.com/nodepool: test-np1
type: h
vmConfig:
extraConfig:
sched.cpu.latencySensitivity: high
revision: "1643191324469087192"
Example after change
CCLI@tph-ci-vmw-caas-mgmt1>>kubectl get vmconfigset -n test test-np1 -o yaml
apiVersion: acm.vmware.com/v1alpha1
kind: VMConfigSet
metadata:
annotations:
kubectl.kubernetes.io/last-applied-configuration: |
{"apiVersion":"acm.vmware.com/v1alpha1","kind":"VMConfigSet","metadata":{"annotations":{},"name":"test-np1","namespace":"test"},"spec":{"nodeLabels":{"telco.vmware.com/nodepool":"test-np1","type":"h"},"vmConfig":{"extraConfig":{"sched.cpu.latencySensitivity":"high"},"preferHosts":null,"revision":"1643191324469087192"}}}
creationTimestamp: "2022-01-24T14:20:12Z"
generation: 3
managedFields:
- apiVersion: acm.vmware.com/v1alpha1
fieldsType: FieldsV1
fieldsV1:
f:metadata:
f:annotations:
.: {}
f:kubectl.kubernetes.io/last-applied-configuration: {}
f:spec:
.: {}
f:nodeLabels:
.: {}
f:telco.vmware.com/nodepool: {}
f:type: {}
f:vmConfig:
.: {}
f:extraConfig:
.: {}
f:sched.cpu.latencySensitivity: {}
f:revision: {}
manager: kubectl-client-side-apply
operation: Update
time: "2022-01-24T15:41:32Z"
name: test-np1
namespace: test
resourceVersion: "82684707"
uid: d821db3b-e062-48d9-9cb5-e61195ec63d6
spec:
nodeLabels:
telco.vmware.com/nodepool: test-np1
type: h
vmConfig:
extraConfig:
sched.cpu.latencySensitivity: high
preferHosts: []