[PR #10227/283d947f backport][stable-11] pacemaker_cluster: enhancements and add unit tests (#10408)

pacemaker_cluster: enhancements and add unit tests (#10227)

* feat(initial): Add unit tests and rewrite pacemaker_cluster

This commit introduces unit tests and pacemaker_cluster module rewrite
to use the pacemaker module utils.

* feat(cleanup): Various fixes and add resource state

This commit migrates the pacemaker_cluster's cleanup state to the
pacemaker_resource module. Additionally, the unit tests for
pacemaker_cluster have been corrected to proper mock run command order.

* doc(botmeta): Add author to pacemaker_cluster

* style(whitespace): Cleanup test files

* refactor(cleanup): Remove unused state value

* bug(fix): Parse apply_all as separate option

* refactor(review): Apply code review suggestions

This commit refactors breaking changes in pacemaker_cluster module into
deprecated features. The following will be scheduled for deprecation:
`state: cleanup` and `state: None`.

* Apply suggestions from code review



* refactor(review): Additional review suggestions

* refactor(deprecations): Remove all deprecation changes

* refactor(review): Enhance rename changelog entry and fix empty string logic

* refactor(cleanup): Remove from pacemaker_resource

* Apply suggestions from code review



* refactor(review): Add changelog and revert required name

* revert(default): Use default state=present

* Update changelogs/fragments/10227-pacemaker-cluster-and-resource-enhancement.yml



* Update changelog fragment.

---------


(cherry picked from commit 283d947)

Co-authored-by: Dexter <45038532+munchtoast@users.noreply.github.com>
Co-authored-by: Felix Fontein <felix@fontein.de>

Author: 3 people

.github/BOTMETA.yml

@@ -1057,7 +1057,7 @@ files:
   $modules/ovh_monthly_billing.py:
     maintainers: fraff
   $modules/pacemaker_cluster.py:
-    maintainers: matbu
+    maintainers: matbu munchtoast
   $modules/pacemaker_resource.py:
     maintainers: munchtoast
   $modules/packet_:

changelogs/fragments/10227-pacemaker-cluster-and-resource-enhancement.yml

@@ -0,0 +1,7 @@
+deprecated_features:
+  - pacemaker_cluster - the parameter ``state`` will become a required parameter in community.general 12.0.0 (https://github.com/ansible-collections/community.general/pull/10227).
+
+minor_changes:
+  - pacemaker_cluster - add ``state=maintenance`` for managing pacemaker maintenance mode (https://github.com/ansible-collections/community.general/issues/10200, https://github.com/ansible-collections/community.general/pull/10227).
+  - pacemaker_cluster - rename ``node`` to ``name`` and add ``node`` alias (https://github.com/ansible-collections/community.general/pull/10227).
+  - pacemaker_resource - enhance module by removing duplicative code (https://github.com/ansible-collections/community.general/pull/10227).

plugins/module_utils/pacemaker.py

@@ -14,7 +14,12 @@
     "absent": "remove",
     "status": "status",
     "enabled": "enable",
-    "disabled": "disable"
+    "disabled": "disable",
+    "online": "start",
+    "offline": "stop",
+    "maintenance": "set",
+    "config": "config",
+    "cleanup": "cleanup",
 }
 
 
@@ -38,27 +43,27 @@ def fmt_resource_argument(value):
 
 
 def get_pacemaker_maintenance_mode(runner):
-    with runner("config") as ctx:
-        rc, out, err = ctx.run()
+    with runner("cli_action config") as ctx:
+        rc, out, err = ctx.run(cli_action="property")
         maintenance_mode_output = list(filter(lambda string: "maintenance-mode=true" in string.lower(), out.splitlines()))
         return bool(maintenance_mode_output)
 
 
-def pacemaker_runner(module, cli_action=None, **kwargs):
+def pacemaker_runner(module, **kwargs):
     runner_command = ['pcs']
-    if cli_action:
-        runner_command.append(cli_action)
     runner = CmdRunner(
         module,
         command=runner_command,
         arg_formats=dict(
+            cli_action=cmd_runner_fmt.as_list(),
             state=cmd_runner_fmt.as_map(_state_map),
             name=cmd_runner_fmt.as_list(),
             resource_type=cmd_runner_fmt.as_func(fmt_resource_type),
             resource_option=cmd_runner_fmt.as_list(),
             resource_operation=cmd_runner_fmt.as_func(fmt_resource_operation),
             resource_meta=cmd_runner_fmt.stack(cmd_runner_fmt.as_opt_val)("meta"),
             resource_argument=cmd_runner_fmt.as_func(fmt_resource_argument),
+            apply_all=cmd_runner_fmt.as_bool("--all"),
             wait=cmd_runner_fmt.as_opt_eq_val("--wait"),
             config=cmd_runner_fmt.as_fixed("config"),
             force=cmd_runner_fmt.as_bool("--force"),

plugins/modules/pacemaker_cluster.py

@@ -13,6 +13,7 @@
 short_description: Manage pacemaker clusters
 author:
   - Mathieu Bultel (@matbu)
+  - Dexter Le (@munchtoast)
 description:
   - This module can manage a pacemaker cluster and nodes from Ansible using the pacemaker CLI.
 extends_documentation_fragment:
@@ -26,18 +27,20 @@
   state:
     description:
       - Indicate desired state of the cluster.
-    choices: [cleanup, offline, online, restart]
+      - The value V(maintenance) has been added in community.general 11.1.0.
+    choices: [cleanup, offline, online, restart, maintenance]
     type: str
-  node:
+  name:
     description:
       - Specify which node of the cluster you want to manage. V(null) == the cluster status itself, V(all) == check the status
         of all nodes.
     type: str
+    aliases: ['node']
   timeout:
     description:
-      - Timeout when the module should considered that the action has failed.
-    default: 300
+      - Timeout period (in seconds) for polling the cluster operation.
     type: int
+    default: 300
   force:
     description:
       - Force the change of the cluster state.
@@ -63,132 +66,104 @@
   returned: always
 """
 
-import time
-
-from ansible.module_utils.basic import AnsibleModule
-
-
-_PCS_CLUSTER_DOWN = "Error: cluster is not currently running on this node"
-
-
-def get_cluster_status(module):
-    cmd = ["pcs", "cluster", "status"]
-    rc, out, err = module.run_command(cmd)
-    if out in _PCS_CLUSTER_DOWN:
-        return 'offline'
-    else:
-        return 'online'
-
-
-def get_node_status(module, node='all'):
-    node_l = ["all"] if node == "all" else []
-    cmd = ["pcs", "cluster", "pcsd-status"] + node_l
-    rc, out, err = module.run_command(cmd)
-    if rc == 1:
-        module.fail_json(msg="Command execution failed.\nCommand: `%s`\nError: %s" % (cmd, err))
-    status = []
-    for o in out.splitlines():
-        status.append(o.split(':'))
-    return status
+from ansible_collections.community.general.plugins.module_utils.module_helper import StateModuleHelper
+from ansible_collections.community.general.plugins.module_utils.pacemaker import pacemaker_runner, get_pacemaker_maintenance_mode
 
 
-def clean_cluster(module, timeout):
-    cmd = ["pcs", "resource", "cleanup"]
-    rc, out, err = module.run_command(cmd)
-    if rc == 1:
-        module.fail_json(msg="Command execution failed.\nCommand: `%s`\nError: %s" % (cmd, err))
-
-
-def set_cluster(module, state, timeout, force):
-    if state == 'online':
-        cmd = ["pcs", "cluster", "start"]
-    if state == 'offline':
-        cmd = ["pcs", "cluster", "stop"]
-        if force:
-            cmd = cmd + ["--force"]
-    rc, out, err = module.run_command(cmd)
-    if rc == 1:
-        module.fail_json(msg="Command execution failed.\nCommand: `%s`\nError: %s" % (cmd, err))
-
-    t = time.time()
-    ready = False
-    while time.time() < t + timeout:
-        cluster_state = get_cluster_status(module)
-        if cluster_state == state:
-            ready = True
-            break
-    if not ready:
-        module.fail_json(msg="Failed to set the state `%s` on the cluster\n" % (state))
+class PacemakerCluster(StateModuleHelper):
+    module = dict(
+        argument_spec=dict(
+            state=dict(type='str', choices=[
+                'cleanup', 'offline', 'online', 'restart', 'maintenance']),
+            name=dict(type='str', aliases=['node']),
+            timeout=dict(type='int', default=300),
+            force=dict(type='bool', default=True)
+        ),
+        supports_check_mode=True,
+    )
+    default_state = ""
+
+    def __init_module__(self):
+        self.runner = pacemaker_runner(self.module)
+        self.vars.set('apply_all', True if not self.module.params['name'] else False)
+        get_args = dict([('cli_action', 'cluster'), ('state', 'status'), ('name', None), ('apply_all', self.vars.apply_all)])
+        if self.module.params['state'] == "maintenance":
+            get_args['cli_action'] = "property"
+            get_args['state'] = "config"
+            get_args['name'] = "maintenance-mode"
+        elif self.module.params['state'] == "cleanup":
+            get_args['cli_action'] = "resource"
+            get_args['name'] = self.module.params['name']
+
+        self.vars.set('get_args', get_args)
+        self.vars.set('previous_value', self._get()['out'])
+        self.vars.set('value', self.vars.previous_value, change=True, diff=True)
+
+        if not self.module.params['state']:
+            self.module.deprecate(
+                'Parameter "state" values not set is being deprecated. Make sure to provide a value for "state"',
+                version='12.0.0',
+                collection_name='community.general'
+            )
+
+    def __quit_module__(self):
+        self.vars.set('value', self._get()['out'])
+
+    def _process_command_output(self, fail_on_err, ignore_err_msg=""):
+        def process(rc, out, err):
+            if fail_on_err and rc != 0 and err and ignore_err_msg not in err:
+                self.do_raise('pcs failed with error (rc={0}): {1}'.format(rc, err))
+            out = out.rstrip()
+            return None if out == "" else out
+        return process
+
+    def _get(self):
+        with self.runner('cli_action state name') as ctx:
+            result = ctx.run(cli_action=self.vars.get_args['cli_action'], state=self.vars.get_args['state'], name=self.vars.get_args['name'])
+            return dict([('rc', result[0]),
+                         ('out', result[1] if result[1] != "" else None),
+                         ('err', result[2])])
+
+    def state_cleanup(self):
+        with self.runner('cli_action state name', output_process=self._process_command_output(True, "Fail"), check_mode_skip=True) as ctx:
+            ctx.run(cli_action='resource')
+
+    def state_offline(self):
+        with self.runner('cli_action state name apply_all wait',
+                         output_process=self._process_command_output(True, "not currently running"),
+                         check_mode_skip=True) as ctx:
+            ctx.run(cli_action='cluster', apply_all=self.vars.apply_all, wait=self.module.params['timeout'])
+
+    def state_online(self):
+        with self.runner('cli_action state name apply_all wait',
+                         output_process=self._process_command_output(True, "currently running"),
+                         check_mode_skip=True) as ctx:
+            ctx.run(cli_action='cluster', apply_all=self.vars.apply_all, wait=self.module.params['timeout'])
+
+        if get_pacemaker_maintenance_mode(self.runner):
+            with self.runner('cli_action state name', output_process=self._process_command_output(True, "Fail"), check_mode_skip=True) as ctx:
+                ctx.run(cli_action='property', state='maintenance', name='maintenance-mode=false')
+
+    def state_maintenance(self):
+        with self.runner('cli_action state name',
+                         output_process=self._process_command_output(True, "Fail"),
+                         check_mode_skip=True) as ctx:
+            ctx.run(cli_action='property', name='maintenance-mode=true')
+
+    def state_restart(self):
+        with self.runner('cli_action state name apply_all wait',
+                         output_process=self._process_command_output(True, "not currently running"),
+                         check_mode_skip=True) as ctx:
+            ctx.run(cli_action='cluster', state='offline', apply_all=self.vars.apply_all, wait=self.module.params['timeout'])
+            ctx.run(cli_action='cluster', state='online', apply_all=self.vars.apply_all, wait=self.module.params['timeout'])
+
+        if get_pacemaker_maintenance_mode(self.runner):
+            with self.runner('cli_action state name', output_process=self._process_command_output(True, "Fail"), check_mode_skip=True) as ctx:
+                ctx.run(cli_action='property', state='maintenance', name='maintenance-mode=false')
 
 
 def main():
-    argument_spec = dict(
-        state=dict(type='str', choices=['online', 'offline', 'restart', 'cleanup']),
-        node=dict(type='str'),
-        timeout=dict(type='int', default=300),
-        force=dict(type='bool', default=True),
-    )
-
-    module = AnsibleModule(
-        argument_spec,
-        supports_check_mode=True,
-    )
-    changed = False
-    state = module.params['state']
-    node = module.params['node']
-    force = module.params['force']
-    timeout = module.params['timeout']
-
-    if state in ['online', 'offline']:
-        # Get cluster status
-        if node is None:
-            cluster_state = get_cluster_status(module)
-            if cluster_state == state:
-                module.exit_json(changed=changed, out=cluster_state)
-            else:
-                if module.check_mode:
-                    module.exit_json(changed=True)
-                set_cluster(module, state, timeout, force)
-                cluster_state = get_cluster_status(module)
-                if cluster_state == state:
-                    module.exit_json(changed=True, out=cluster_state)
-                else:
-                    module.fail_json(msg="Fail to bring the cluster %s" % state)
-        else:
-            cluster_state = get_node_status(module, node)
-            # Check cluster state
-            for node_state in cluster_state:
-                if node_state[1].strip().lower() == state:
-                    module.exit_json(changed=changed, out=cluster_state)
-                else:
-                    if module.check_mode:
-                        module.exit_json(changed=True)
-                    # Set cluster status if needed
-                    set_cluster(module, state, timeout, force)
-                    cluster_state = get_node_status(module, node)
-                    module.exit_json(changed=True, out=cluster_state)
-
-    elif state == 'restart':
-        if module.check_mode:
-            module.exit_json(changed=True)
-        set_cluster(module, 'offline', timeout, force)
-        cluster_state = get_cluster_status(module)
-        if cluster_state == 'offline':
-            set_cluster(module, 'online', timeout, force)
-            cluster_state = get_cluster_status(module)
-            if cluster_state == 'online':
-                module.exit_json(changed=True, out=cluster_state)
-            else:
-                module.fail_json(msg="Failed during the restart of the cluster, the cluster cannot be started")
-        else:
-            module.fail_json(msg="Failed during the restart of the cluster, the cluster cannot be stopped")
-
-    elif state == 'cleanup':
-        if module.check_mode:
-            module.exit_json(changed=True)
-        clean_cluster(module, timeout)
-        cluster_state = get_cluster_status(module)
-        module.exit_json(changed=True, out=cluster_state)
+    PacemakerCluster.execute()
 
 
 if __name__ == '__main__':