Merge pull request #157 from meraki/develop

v2.25.1

Author: fmunozmiranda

.github/workflows/ansible-lint.yml

@@ -10,7 +10,7 @@ jobs:
     steps:
       - uses: actions/checkout@v4
       - name: Run ansible-lint
-        uses: ansible/ansible-lint@main
+        uses: ansible/ansible-lint@v26
         with:
           args: ""
           setup_python: "true"

.github/workflows/certification.yml

@@ -35,15 +35,16 @@ concurrency:
 # for each affected version of ansible-core (for example, `tests/sanity/ignore-2.18.txt`) and add corresponding entries.
 jobs:
   call:
-    uses: ansible-collections/partner-certification-checker/.github/workflows/certification-reusable.yml@v2.0.0 # zizmor: ignore[unpinned-uses]
+    uses: ansible-collections/partner-certification-checker/.github/workflows/certification-reusable.yml@v4.1.0 # zizmor: ignore[unpinned-uses]
     # If the minimal Ansible Core version your collection supports Ansible
     # is greater than 2.16.0, specify it below.
     # You might also have to specify Python version supported by that version,
     # see https://docs.ansible.com/projects/ansible/latest/reference_appendices/release_and_maintenance.html#ansible-core-support-matrix
     # with:
     #   ansible-core-version: '2.17.0'
-    #   python-versions: '["3.7", "3.8", "3.9", "3.10", "3.11", "3.12"]'
+    #   origin-python-version: '3.12'
     #   collection-root: 'ansible_collections/junipernetworks/junos'
+    #   collection-deps: ''
     #   automation-hub: true
     # Uncomment the lines below if you have 'automation-hub: true' and need to install collections from Red Hat Ansible Automation Hub.
     # This requires setting a repository secret called 'AH_TOKEN' which value contains a Red Hat Ansible Automation Hub token.

CHANGELOG.rst

@@ -4,6 +4,20 @@ Cisco.Meraki Release Notes
 
 .. contents:: Topics
 
+v2.25.1
+=======
+
+Bugfixes
+--------
+
+- devices_management_interface - Removed the erroneous mapping of the create operation to rebootDevice; state=present now always updates the management
+  interface, since it is a resource that always pre-exists and has no create operation.
+- devices_management_interface - Removed the bare exception handler in get_object_by_name() that swallowed API read failures (including the module's own
+  fail_json) and misreported them as the object not existing.
+- plugin_utils/meraki - exec_meraki() now fails the task when a write operation's API response explicitly reports "success": false (as rebootDevice and
+  similar fire-and-forget endpoints do on rejection), instead of letting the calling module report changed=true/failed=false. Responses without a "success"
+  key, and read operations, are unaffected.
+
 v2.25.0
 =======
 

README.md

@@ -160,7 +160,7 @@ This collection has been tested against the following environments:
 | 1.57.0                     | 2.21.2                 | 1.57.0              |
 | 1.68.0                     | 2.23.0                 | 2.2.0               |
 | 1.69.0                     | 2.24.2                 | 3.0.0               |
-| 1.72.0                     | 2.25.0                 | 4.4.0               |
+| 1.72.0                     | 2.25.1                 | 4.4.0               |
 
 *Notes*:
 

changelogs/changelog.yaml

@@ -1640,3 +1640,14 @@ releases:
         - "devices_wireless_bluetooth_settings - Fixed a malformed UUID in the module's EXAMPLES block."
         - "networks_wireless_bluetooth_settings - Fixed a malformed UUID in the module's EXAMPLES block."
     release_date: "2026-08-10"
+  2.25.1:
+    changes:
+      bugfixes:
+        - "devices_management_interface - Removed the erroneous mapping of the create operation to rebootDevice; state=present now always updates the management
+          interface, since it is a resource that always pre-exists and has no create operation."
+        - "devices_management_interface - Removed the bare exception handler in get_object_by_name() that swallowed API read failures (including the module's
+          own fail_json) and misreported them as the object not existing."
+        - "plugin_utils/meraki - exec_meraki() now fails the task when a write operation's API response explicitly reports \"success\": false (as rebootDevice
+          and similar fire-and-forget endpoints do on rejection), instead of letting the calling module report changed=true/failed=false. Responses without a
+          \"success\" key, and read operations, are unaffected."
+    release_date: "2026-08-18"

galaxy.yml

@@ -1,7 +1,7 @@
 ---
 namespace: cisco
 name: meraki
-version: 2.25.0
+version: 2.25.1
 readme: README.md
 authors:
   - Francisco Muñoz <fmunoz@cloverhound.com>

plugins/action/devices_management_interface.py

@@ -60,14 +60,6 @@ def get_all_params(self, name=None, id=None):
             new_object_params['serial'] = self.new_object.get('serial')
         return new_object_params
 
-    def create_params(self):
-        new_object_params = {}
-        if self.new_object.get('serial') is not None or self.new_object.get(
-                'serial') is not None:
-            new_object_params['serial'] = self.new_object.get('serial') or \
-                self.new_object.get('serial')
-        return new_object_params
-
     def update_all_params(self):
         new_object_params = {}
         if self.new_object.get('wan1') is not None or self.new_object.get(
@@ -87,21 +79,20 @@ def update_all_params(self):
     def get_object_by_name(self, name):
         result = None
         # NOTE: Does not have a get by name method, using get all
-        try:
-            items = self.meraki.exec_meraki(
-                family="devices",
-                function="getDeviceManagementInterface",
-                params=self.get_all_params(name=name),
-            )
-            if isinstance(items, dict):
-                if 'response' in items:
-                    items = items.get('response')
-            result = get_dict_result(items, 'name', name)
-            if result is None:
-                result = items
-        except Exception as e:
-            print("Error: ", e)
-            result = None
+        # exec_meraki() already raises AnsibleActionFail on an API error, so a
+        # failed read (bad serial, rate limit, etc.) fails the task instead of
+        # being mistaken for "the object does not exist".
+        items = self.meraki.exec_meraki(
+            family="devices",
+            function="getDeviceManagementInterface",
+            params=self.get_all_params(name=name),
+        )
+        if isinstance(items, dict):
+            if 'response' in items:
+                items = items.get('response')
+        result = get_dict_result(items, 'name', name)
+        if result is None:
+            result = items
         return result
 
     def get_object_by_id(self, id):
@@ -147,15 +138,6 @@ def requires_update(self, current_obj):
                 meraki_param,
                 ansible_param) in obj_params)
 
-    def create(self):
-        result = self.meraki.exec_meraki(
-            family="devices",
-            function="rebootDevice",
-            params=self.create_params(),
-            op_modifies=True,
-        )
-        return result
-
     def update(self):
         id = self.new_object.get("id")
         name = self.new_object.get("name")
@@ -210,17 +192,22 @@ def run(self, tmp=None, task_vars=None):
 
         response = None
         if state == "present":
+            # A device's management interface always exists; there is no
+            # create operation for it (the only write is a PUT). Reaching
+            # this point with obj_exists False means the read returned an
+            # unexpected shape rather than failing outright, since a real
+            # failure already raised inside exists() -> get_object_by_name().
             (obj_exists, prev_obj) = obj.exists()
-            if obj_exists:
-                if obj.requires_update(prev_obj):
-                    response = obj.update()
-                    meraki.object_updated()
-                else:
-                    response = prev_obj
-                    meraki.object_already_present()
+            if not obj_exists:
+                raise AnsibleActionFail(
+                    "Unable to retrieve the current management interface for serial '{serial}'.".format(
+                        serial=self._task.args.get("serial")))
+            if obj.requires_update(prev_obj):
+                response = obj.update()
+                meraki.object_updated()
             else:
-                response = obj.create()
-                meraki.object_created()
+                response = prev_obj
+                meraki.object_already_present()
 
         self._result.update(dict(meraki_response=response))
         self._result.update(meraki.exit_json())

plugins/modules/devices_management_interface.py

@@ -9,10 +9,10 @@
 module: devices_management_interface
 short_description: Resource module for devices _management _interface
 description:
-  - Manage operations create and update of the resource devices _management _interface.
-  - Reboot a device. This endpoint has a sustained rate limit of one request every
-    60 seconds.
+  - Manage operations update of the resource devices _management _interface.
   - Update the management interface settings for a device.
+  - A device's management interface always exists, so C(state=present) always
+    updates it; there is no create operation for this resource.
 version_added: '1.0.0'
 extends_documentation_fragment:
   - cisco.meraki.module
@@ -81,18 +81,13 @@
   - meraki >= 2.4.9
   - python >= 3.5
 seealso:
-  - name: Cisco Meraki documentation for devices rebootDevice
-    description: Complete reference of the rebootDevice API.
-    link: https://developer.cisco.com/meraki/api-v1/#!reboot-device
   - name: Cisco Meraki documentation for devices updateDeviceManagementInterface
     description: Complete reference of the updateDeviceManagementInterface API.
     link: https://developer.cisco.com/meraki/api-v1/#!update-device-management-interface
 notes:
-  - SDK Method used are
-    devices.Devices.reboot_device,
+  - SDK Method used is
     devices.Devices.update_device_management_interface,
   - Paths used are
-    post /devices/{serial}/reboot,
     put /devices/{serial}/managementInterface,
 """
 
@@ -142,31 +137,6 @@
       usingStaticIp: false
       vlan: 2
       wanEnabled: enabled
-- name: Create
-  cisco.meraki.devices_management_interface:
-    meraki_api_key: "{{ meraki_api_key }}"
-    meraki_base_url: "{{ meraki_base_url }}"
-    meraki_single_request_timeout: "{{ meraki_single_request_timeout }}"
-    meraki_certificate_path: "{{ meraki_certificate_path }}"
-    meraki_requests_proxy: "{{ meraki_requests_proxy }}"
-    meraki_wait_on_rate_limit: "{{ meraki_wait_on_rate_limit }}"
-    meraki_nginx_429_retry_wait_time: "{{ meraki_nginx_429_retry_wait_time }}"
-    meraki_action_batch_retry_wait_time: "{{ meraki_action_batch_retry_wait_time }}"
-    meraki_retry_4xx_error: "{{ meraki_retry_4xx_error }}"
-    meraki_retry_4xx_error_wait_time: "{{ meraki_retry_4xx_error_wait_time }}"
-    meraki_maximum_retries: "{{ meraki_maximum_retries }}"
-    meraki_output_log: "{{ meraki_output_log }}"
-    meraki_log_file_prefix: "{{ meraki_log_file_prefix }}"
-    meraki_log_path: "{{ meraki_log_path }}"
-    meraki_print_console: "{{ meraki_print_console }}"
-    meraki_suppress_logging: "{{ meraki_suppress_logging }}"
-    meraki_simulate: "{{ meraki_simulate }}"
-    meraki_be_geo_id: "{{ meraki_be_geo_id }}"
-    meraki_caller: "{{ meraki_caller }}"
-    meraki_use_iterator_for_get_pages: "{{ meraki_use_iterator_for_get_pages }}"
-    meraki_inherit_logging_config: "{{ meraki_inherit_logging_config }}"
-    state: present
-    serial: string
 """
 RETURN = r"""
 meraki_response:

plugins/plugin_utils/meraki.py

@@ -285,7 +285,7 @@ def __init__(self, params):
                 suppress_logging=params.get("meraki_suppress_logging"),
                 simulate=params.get("meraki_simulate"),
                 be_geo_id=params.get("meraki_be_geo_id"),
-                caller="MerakiAnsibleCollection/2.25.0 Cisco",
+                caller="MerakiAnsibleCollection/2.25.1 Cisco",
                 use_iterator_for_get_pages=params.get(
                     "meraki_use_iterator_for_get_pages"),
                 inherit_logging_config=params.get(
@@ -334,6 +334,7 @@ def extract_file_name(self, file_path):
         return os.path.basename(file_path)
 
     def exec_meraki(self, family, function, params=None, op_modifies=False, **kwargs):
+        family_name = family
         try:
             family = getattr(self.api, family)
             func = getattr(family, function)
@@ -369,6 +370,22 @@ def exec_meraki(self, family, function, params=None, op_modifies=False, **kwargs
                     "The error was: {error}"
                 ).format(error=to_native(e))
             )
+        # Some write operations (reboot, cycle ports, blink LEDs, unenroll,
+        # batch updates, ...) return 2xx with a body of the form
+        # {"success": false} instead of raising an APIError when the
+        # requested action is rejected. Without this check that response
+        # still looks like a normal successful result to the calling action
+        # plugin, which reports changed=True/failed=False even though
+        # nothing happened. A response with no "success" key (e.g. the
+        # updated object itself, as returned by most update* operations) is
+        # unaffected: only an explicit `False` is treated as a failure.
+        if op_modifies and isinstance(response, dict) and response.get("success") is False:
+            self.fail_json(
+                msg=(
+                    "{family}, {function} - the operation was not successful "
+                    "(the API response reported success: false). Response: {response}"
+                ).format(family=family_name, function=function, response=response)
+            )
         return response
 
     def fail_json(self, msg, **kwargs):