meraki_device making wrong API call

[victoreebel]

Trying the first example from the doc 'query all devices', substituting the dummy values for my own:

    - name: Query all devices in a network.
      cisco.meraki.meraki_device:
        auth_key: MYKEY
        org_name: MYORG
        net_name: MYNET
        state: query
      delegate_to: localhost

Might be missing something, but I'm getting this return -

"msg": 404,
"response": "OK (unknown bytes)",
"status": 404,
"url": "https://api.meraki.com/api/v1/organizations/<MY ORG ID>/configTemplates" 

So it doesn't seem to be hitting the devices API, rather the configTemplates one.

Module version is 2.15.3

[kbreit]

Thank you for reporting this. It's quite an interesting bug as well. I do store some URLs in the shared code between modules but I can't tell why that particular URL is called. Give me a little time to look into this.

[kbreit]

I think I've identified why it's failing. Can you try to use net_id instead of net_name? I expect it will work. If it does work, please let me know if your organization has any config templates.

[kbreit]

My theory was if an organization had no configuration templates it would give a 404 but that's not the case. Unfortunately, I can't reproduce the bug. Can you describe your organization/network environment in some detail, especially around the configuration templates, so I can see if I can reproduce it? Thank you.

[victoreebel]

Thanks for this - I've tried using every combination of org_name, org_id, net_name and net_id with two networks, but get the same response in all cases (the org string being passed to the config_templates call).

I also encountered a similar issue with the ACLs module, and in that case was actually able to get that to work by using the net_id rather than net_name, interestingly.

The org in question is currently in the process of being handed over to us, so don't have full visibility r.e. configuration templates, but will try to find out.

[kbreit]

Thank you. Let me know what you find out. I haven't been able to reproduce it but I think I see where it's happening. When the module downloads all the networks it also downloads all the configuration templates at https://github.com/CiscoDevNet/ansible-meraki/blob/6456adf4f82dcb7931054bf411a270014f5baaf4/plugins/module_utils/network/meraki/meraki.py#L267. Eventually that makes a call to my request() method and if it has an error code above 400, it errors https://github.com/CiscoDevNet/ansible-meraki/blob/6456adf4f82dcb7931054bf411a270014f5baaf4/plugins/module_utils/network/meraki/meraki.py#L436C13-L436C13.

By chance, do you have access to an API tool like Postman to do some testing to see how we can get that to work?

[victoreebel]

Thanks for looking into this. Just heard from the other team, they confirmed that no configuration templates are currently in use.

Also tried omitting both net_id and net_name from the play (so ommitting network altogether) and just using org_name, and still hitting the same error.

I also encountered a similar issue with the ACLs module, and in that case was actually able to get that to work by using the net_id rather than net_name, interestingly.

This I've now noticed with several modules, but in all other cases except the meraki_device one, using net_id prevents the issue from happening.

The organization currently consists of 3 distinct networks - an appliance one, switched and wireless and a combined one.

I do have Postman setup. Which tests did you have in mind? Right off the bat I can say that running {{baseUrl}}/organizations/:organizationId/devices returns succesfully.

[kbreit]

Can you try testing /organizations/{organizationId}/configTemplates? I want to see what happens with that. Also, please try commenting out https://github.com/CiscoDevNet/ansible-meraki/blob/6456adf4f82dcb7931054bf411a270014f5baaf4/plugins/module_utils/network/meraki/meraki.py#L267-L269 from the file specified in the link. That's where I think it's doing the lookup so if commenting out that code fixes the problem, it at least isolates it.