Issue re-managing a node - New-style module did not handle its own exit

[Andyjb8]

Thanks you for this collection! I am testing it out as this will be very useful and the solarwinds.orion.orion_node module works great to add nodes and to set a managed node to unmanaged, but I am having an issue setting an un-managed state back to a managed state.

Here is my task:

• name: RE-MANANAGE NODE in Solarwinds solarwinds.orion.orion_node: <<: *solarwinds_info state: managed # unmanaged name: "{{ inventory_hostname }}" ip_address: "{{ ansible_host }}" snmp_version: 3 snmpv3_auth_key: xxxxxxxxx snmpv3_auth_method: SHA1 snmpv3_priv_key: xxxxxxxxx snmpv3_priv_method: AES128 snmpv3_username: xxxxxxxxx delegate_to: localhost

Here is the error I get: fatal: [myhostname -> localhost]: FAILED! => { "changed": false, "msg": "New-style module did not handle its own exit"

Also, I found it necessary to specify the snmp_version and parameters to manage / un-manage a node. Why is that necessary?

Thanks!

[jeisenbath]

Hey! I really appreciate you testing, I can see right away the issue with setting a node back to a managed state (and this might be a good time to add a CHANGELOG actually).

To the other point of requiring snmp_version and parameters, that should not be necessary. However if you do specify snmp_version, it will require the additional params for that version, and I should see about updating the documentation block to reflect that. I use both unmanaged and muted states for orion_node without needing to pass snmp_verison, can you retest something like this for me?

- name: Unmanage a node
  solarwinds.orion.orion_node:
    hostname: "{{ solarwinds_host }}"
    username: "{{ solarwinds_username }}"
    password: "{{ solarwinds_password }}"
    name: "{{ node_name }}"
    state: unmanaged

[Andyjb8]

Hmm, I am getting:

"msg": "snmp_version is 2 but all of the following are missing: ro_community_string"

Here is what I have:

- hosts: all
  gather_facts: false

  vars:
    solarwinds_info: &solarwinds_info
      hostname: al-svr-sol-01
      username: user
      password: password

  tasks:  
- name: Unmanage Node Solarwinds
    solarwinds.orion.orion_node:
      <<: *solarwinds_info
      state: unmanaged
      name: "{{ inventory_hostname }}"
      ip_address: "{{ ansible_host }}"

Originally ip_address in there, I took that out and I get the same thing.

- name: Unmanage Node Solarwinds
    solarwinds.orion.orion_node:
      <<: *solarwinds_info
      state: unmanaged
      name: "{{ inventory_hostname }}"
      ip_address: "{{ ansible_host }}"

[Andyjb8]

I just realized what I had above was changing state to unmanaged. I get the same "msg": "snmp_version is 2 but all of the following are missing: ro_community_string" warning when changing state to either managed or unmanaged. But, when I have the snmpv3 parameters unmanaged works, but managed does not.

[jeisenbath]

awesome yup I see my mistake there as well. I am setting a default for snmp version when it isn't passed, which is triggering the required_if on that parameter. I have a few changes to make but should have a new version by tomorrow morning.

[Andyjb8]

Excellent, thanks!

Also, should I be able to update snmpv3 credentials like this with orion_update_node?

  - name: Update SNMPv3 on Node
    solarwinds.orion.orion_update_node:
      <<: *solarwinds_info
      name: "{{ inventory_hostname }}"
      properties:
        ObjectSubType: SNMP
        snmp_version: 3
        snmpv3_auth_key: authpass
        snmpv3_auth_method: SHA1
        snmpv3_priv_key: privpass
        snmpv3_priv_method: AES128
        snmpv3_username: snmpuser
    delegate_to: localhost

It's not working and I get this message:

_"msg": "Failed to update 400 Client Error: Invalid column name 'snmp_version'.\r\nInvalid column name 'snmpv3_auth_key'.\r\nInvalid column name 'snmpv3_auth_method'.\r\nInvalid column name 'snmpv3_priv_key'.\r\nInvalid column name 'snmpv3_priv_method'.\r\nInvalid column name 'snmpv3_username'. for url: https://x.x.x.x:17778/SolarWinds/InformationService/v3/Json/swis:/x.x.x.x/Orion/Orion.Nodes/NodeID=1061"_

[jeisenbath]

solarwinds treats snmpv3 credentials differently, they are stored in a credential library, so that may be the issue that those credentials aren't updated by the node's URI. This is a separate issue that I haven't had the need to correct, but managing nodes with snmpv3 this way will absolutely fill up your Credential Library in solarwinds (it will make a new credential set for each node)