Fix GPU Operator node prep for CentOS #868
Description
Running the k8s-cluster.yml with th GPU operator flag to true on a fresh CentOS 7 install results in the following error. This is currently failing in the nightly tests.
TASK [nvidia-gpu-operator-node-prep : load required kernel modules] ************
changed: [virtual-mgmt01-3] => (item=i2c_core) => changed=true
ansible_loop_var: item
item: i2c_core
name: i2c_core
params: ''
state: present
changed: [virtual-gpu01-3] => (item=i2c_core) => changed=true
ansible_loop_var: item
item: i2c_core
name: i2c_core
params: ''
state: present
changed: [virtual-mgmt01-3] => (item=ipmi_msghandler) => changed=true
ansible_loop_var: item
item: ipmi_msghandler
name: ipmi_msghandler
params: ''
state: present
changed: [virtual-gpu01-3] => (item=ipmi_msghandler) => changed=true
ansible_loop_var: item
item: ipmi_msghandler
name: ipmi_msghandler
params: ''
state: present
RUNNING HANDLER [nvidia-gpu-operator-node-prep : update-initramfs] *************
fatal: [virtual-mgmt01-3]: FAILED! => changed=false
cmd: update-initramfs -u
msg: '[Errno 2] No such file or directory'
rc: 2
fatal: [virtual-gpu01-3]: FAILED! => changed=false
cmd: update-initramfs -u
msg: '[Errno 2] No such file or directory'
rc: 2