community.vmware.vmware_vmkernel

[KWG66]

As a VMware veteran, I find little use in the design of this module to automatically migrate a vmkernel to vDS or to vSS if present. Its a needle in a haystack requirement yet its been baked into this module. WOW...

So, I have developed a playbook that adds a VMware host to vCenter then fully configures it for our environment, start to finish. The community.vmware.vmware_vmkernel is part of the process of course, and it works great.

However, when you run the playbook a second time to verify \ validate the configuration, because the vmkernel is already there it attempts to migrate it and fails, thereby not providing the verification you are looking for.

I find working with Ansible playbooks fairly easy, however, as a newbie I am struggling to figure out how to code my playbook around this module so that if the vmkernel is already present it will skip trying to migrate it, then continue performing the remaining tasks.

Thanks in advance for suggestions

[mariolenz]

If you don't show us your playbook / task and tell us what ansible-core and community.vmware you're running, I think we should close this. Just saying "I'm trying foo but it doesn't work" is not enough.

Apart from this, I'm not really sure if this is really serious : "I find little use in the design of this module to automatically migrate a vmkernel to vDS or to vSS if present. Its a needle in a haystack requirement yet its been baked into this module. WOW"

I'm sorry, but this sounds like you're only ranting. Give us some more information or I'll close this issue.

[KWG66]

Perhaps some rant, questioning the why 'the migration of a vmkernel between switch types' would be a default behavior, but that said. the module is in the title: community.vmware.vmware_vmkernel

It works great, as stated, allows me to accomplish the goal of configuring a host just out of installation.

The problem is as I see it is an idempotency issue. When you run it again as part of a verification the default behavior will cause it to try to migrate the vmk, not desirable. Just looking for ideas on how to NOT migrate the vmk when present so when the playbook runs to simply verify the config.

Here is the task from the playbook, as requested, any ideas on how to configure this so that if the vmk is present already it just moves on provides a nice green text in the playback?

• name: add nfs-X vmk
  community.vmware.vmware_vmkernel:
  hostname: "{{ hostname }}"
  username: "{{ username }}"
  password: "{{ password }}"
  esxi_hostname: "{{ esxi_hostname }}"
  vswitch_name: vSwitch4
  portgroup_name: NFS-X
  mtu: 9000
  network:
  type: 'static'
  ip_address: x.x.x.x
  subnet_mask: 255.255.255.0
  tcpip_stack: default
  state: present
  enable_mgmt: false
  delegate_to: localhost