Improve the agent_files_deletion test

tests/system/README.md

@@ -191,6 +191,15 @@ in `system/provisioning/<specified_cluster_environment>` path:
 ansible-playbook -i inventory.yml playbook.yml
 ```
 
+If you want to specify a particular branch of the Wazuh repository, you will need to include:
+```shell script
+ansible-playbook -i inventory.yml playbook.yml --extra-vars='{"wazuh_branch":"v4.3.0-rc1"}'
+```
+In the **basic cluster**, you also have to specify a branch from the Wazuh QA repository.
+```shell script
+ansible-playbook -i inventory.yml playbook.yml --extra-vars='{"wazuh_branch":"v4.3.0-rc1", "wazuh_qa_branch":"master"}'
+```
+
 We use [pytest](https://docs.pytest.org/en/latest/contents.html) to run our cluster system tests. Pytest will
 recursively look for the closest `conftest` to import all the variables and fixtures needed for every test. If something
 is lacking from the closest one, it will look for the next one (if possible) until reaching the current directory. This

tests/system/provisioning/basic_cluster/roles/master-role/tasks/main.yml

@@ -28,6 +28,13 @@
     dest: /wazuh
     version: "{{ wazuh_branch }}"
 
+- name: "Clone wazuh-qa repository"
+  git:
+    repo: "https://github.com/wazuh/wazuh-qa"
+    dest: /wazuh-qa
+    version: "{{ wazuh_qa_branch }}"
+    depth: '1'
+
 - name: Install master
   args:
     chdir: /wazuh
@@ -85,6 +92,7 @@
       authd.debug=2
       wazuh_clusterd.debug=2
       wazuh_db.debug=2
+      wazuh_modules.debug=2
 
 - name: Register agents
   blockinfile:
@@ -100,3 +108,6 @@
 
 - name: Start Wazuh
   command: /var/ossec/bin/wazuh-control restart
+
+- name: "Install necessary dependencies"
+  command: /var/ossec/framework/python/bin/python3.9 -m pip install lockfile filetype certifi testinfra

tests/system/provisioning/basic_cluster/roles/worker-role/tasks/main.yml

@@ -29,6 +29,13 @@
     dest: /wazuh
     version: "{{ wazuh_branch }}"
 
+- name: "Clone wazuh-qa repository"
+  git:
+    repo: "https://github.com/wazuh/wazuh-qa"
+    dest: /wazuh-qa
+    version: "{{ wazuh_qa_branch }}"
+    depth: '1'
+
 - name: Install worker
   args:
     chdir: /wazuh
@@ -85,6 +92,10 @@
       authd.debug=2
       wazuh_clusterd.debug=2
       wazuh_db.debug=2
+      wazuh_modules.debug=2
 
 - name: Restart Wazuh
   command: /var/ossec/bin/wazuh-control restart
+
+- name: "Install necessary dependencies"
+  command: /var/ossec/framework/python/bin/python3.9 -m pip install lockfile filetype certifi testinfra

tests/system/test_cluster/test_agent_files_deletion/data/get_wdb_agent.py

@@ -0,0 +1,10 @@
+# Copyright (C) 2015-2021, Wazuh Inc.
+# Created by Wazuh, Inc. <info@wazuh.com>.
+# This program is free software; you can redistribute it and/or modify it under the terms of GPLv2
+import sys
+sys.path.append('/wazuh-qa/deps/wazuh_testing')
+from wazuh_testing import wazuh_db
+
+result = wazuh_db.query_wdb(sys.argv[1])
+if result:
+  print(result)

tests/system/test_cluster/test_agent_files_deletion/data/messages.yml

@@ -0,0 +1,19 @@
+---
+wazuh-worker2:
+  - regex: '.*Received.*shared files to update from .*'
+    path: "/var/ossec/logs/cluster.log"
+    timeout: 60
+  - regex: '.*Worker wazuh-worker2.*Integrity sync.*Finished in.*'
+    path: "/var/ossec/logs/cluster.log"
+    timeout: 60
+  - regex: '.*Worker wazuh-worker2.*Integrity sync.*Starting.*'
+    path: "/var/ossec/logs/cluster.log"
+    timeout: 60
+  - regex: '.*Worker wazuh-worker2.*Integrity sync.*Finished in.*'
+    path: "/var/ossec/logs/cluster.log"
+    timeout: 60
+
+wazuh-master:
+  - regex: '.*Worker wazuh-worker2.*Integrity check.*Finished in.*'
+    path: "/var/ossec/logs/cluster.log"
+    timeout: 60

tests/system/test_cluster/test_agent_files_deletion/test_agent_files_deletion.py

@@ -1,46 +1,81 @@
 # Copyright (C) 2015-2021, Wazuh Inc.
 # Created by Wazuh, Inc. <info@wazuh.com>.
 # This program is free software; you can redistribute it and/or modify it under the terms of GPLv2
-
+import os
+import re
 from os.path import join, dirname, abspath
-from time import sleep
+from time import time, sleep
 
 import pytest
-
-from wazuh_testing.tools import WAZUH_PATH
+from wazuh_testing.tools import WAZUH_PATH, WAZUH_LOGS_PATH
+from wazuh_testing.tools.monitoring import HostMonitor
 from wazuh_testing.tools.system import HostManager
 
 master_host = 'wazuh-master'
 worker_host = 'wazuh-worker2'
 agent_host = 'wazuh-agent3'
+local_path = os.path.dirname(os.path.abspath(__file__))
+messages_path = os.path.join(local_path, 'data/messages.yml')
+script_path = os.path.join(re.sub(r'^.*?wazuh-qa', '/wazuh-qa', local_path), 'data/get_wdb_agent.py')
+
+tmp_path = os.path.join(local_path, 'tmp')
 managers_hosts = [master_host, worker_host]
 inventory_path = join(dirname(dirname(dirname(abspath(__file__)))), 'provisioning', 'basic_cluster', 'inventory.yml')
 host_manager = HostManager(inventory_path)
-time_to_sync = 60
+while_time = 5
+time_to_sync = 20
+time_to_agent_reconnect = 180
 
 # Each file should exist in all hosts specified in 'hosts'.
 files = [{'path': join(WAZUH_PATH, 'queue', 'rids', '{id}'), 'hosts': managers_hosts},
          {'path': join(WAZUH_PATH, 'queue', 'agent-groups', '{id}'), 'hosts': managers_hosts},
          {'path': join(WAZUH_PATH, 'queue', 'diff', '{name}'), 'hosts': [worker_host]},
          {'path': join(WAZUH_PATH, 'queue', 'db', '{id}.db'), 'hosts': [worker_host]}]
-db_queries = ["select * from agent where id={id}",
-              "select * from belongs where id_agent={id}"]
+
+queries = ['global sql select * from agent where id={id}',
+           'global sql select * from belongs where id_agent={id}']
 
 
-@pytest.fixture(scope='function')
-def register_agent():
-    """Restart the removed agent to trigger auto-enrollment."""
-    yield
-    host_manager.get_host(agent_host).ansible('command', f'service wazuh-agent restart', check=False)
+def agent_healthcheck(master_token):
+    """Check if the agent is active and reporting."""
+    timeout = time() + time_to_agent_reconnect
+    healthy = False
 
+    while not healthy:
+        response = host_manager.make_api_call(host=master_host, method='GET', token=master_token,
+                                              endpoint='/agents?status=active')
 
-def test_agent_files_deletion(register_agent):
+        assert response['status'] == 200, 'Failed when trying to get the active agents'
+        if int(response['json']['data']['total_affected_items']) == 4:
+            for item in response['json']['data']['affected_items']:
+                if item['name'] == agent_host and item['manager'] == worker_host:
+                    healthy = True
+        elif time() > timeout:
+            print("The agent 'wazuh-agent3' is not 'Active' yet.")
+            break
+        sleep(while_time)
+    sleep(time_to_sync)
+
+
+def test_agent_files_deletion():
     """Check that when an agent is deleted, all its related files in managers are also removed."""
-    # Get the current ID and name of the agent that is reporting to worker_host.
+    # Clean ossec.log and cluster.log
+    for hosts in managers_hosts:
+        host_manager.clear_file(host=hosts, file_path=os.path.join(WAZUH_LOGS_PATH, 'ossec.log'))
+        host_manager.clear_file(host=hosts, file_path=os.path.join(WAZUH_LOGS_PATH, 'cluster.log'))
+        host_manager.control_service(host=hosts, service='wazuh', state='restarted')
+
+    # Get the token
     master_token = host_manager.get_api_token(master_host)
+
+    # Check if the agent is connected and reporting
+    agent_healthcheck(master_token)
+
+    # Get the current ID and name of the agent that is reporting to worker_host.
     response = host_manager.make_api_call(host=master_host, method='GET', token=master_token,
-                                          endpoint=f'/agents?select=id,name&q=manager={worker_host}')
-    assert response['status'] == 200, f'Failed when trying to obtain agent ID: {response}'
+                                          endpoint=f"/agents?select=id,name&q=manager={worker_host}")
+
+    assert response['status'] == 200, f"Failed when trying to obtain agent ID: {response}"
     try:
         agent_id = response['json']['data']['affected_items'][0]['id']
         agent_name = response['json']['data']['affected_items'][0]['name']
@@ -51,43 +86,44 @@ def test_agent_files_deletion(register_agent):
     for file in files:
         for host in file['hosts']:
             result = host_manager.run_shell(
-                host, f'test -e {file["path"].format(id=agent_id, name=agent_name)} && echo "exists"'
+                host, f"test -e {file['path'].format(id=agent_id, name=agent_name)} && echo 'exists'"
             )
-            assert result, f'This file should exist in {host} but could not be found: ' \
-                           f'{file["path"].format(id=agent_id, name=agent_name)}'
+            assert result, f"This file should exist in {host} but could not be found: " \
+                           f"{file['path'].format(id=agent_id, name=agent_name)}"
 
-    # Check that agent information exists in global.db
+    # Check that agent information is in the wdb socket
     for host in managers_hosts:
-        for query in db_queries:
-            result = host_manager.run_command(
-                host,
-                f'sqlite3 {join(WAZUH_PATH, "queue", "db", "global.db")} "{query.format(id=agent_id, name=agent_name)}"'
-            )
-            assert result, f'This db query should have returned something in {host}, but it did not: ' \
-                           f'{query.format(id=agent_id, name=agent_name)}'
+        for query in queries:
+            result = host_manager.run_command(host,
+                                              f"{WAZUH_PATH}/framework/python/bin/python3.9 "
+                                              f"{script_path} '{query.format(id=agent_id)}'")
+            assert result, f"This db query should have returned something in {host}, but it did not: {result}"
 
+    # Remove the agent
     response = host_manager.make_api_call(host=master_host, method='DELETE', token=master_token,
-                                          endpoint=f'/agents?agents_list={agent_id}&status=all&older_than=0s')
-    assert response['status'] == 200, f'Failed when trying to remove agent {agent_id}: {response}'
+                                          endpoint=f"/agents?agents_list={agent_id}&status=all&older_than=0s")
+    assert response['status'] == 200, f"Failed when trying to remove agent {agent_id}: {response}"
 
     # Wait until information is synced to all workers
+    HostMonitor(inventory_path=inventory_path, messages_path=messages_path, tmp_path=tmp_path).run()
     sleep(time_to_sync)
 
     # Check that agent-related files where removed from each node.
     for file in files:
         for host in file['hosts']:
             result = host_manager.run_shell(
-                host, f'test -e {file["path"].format(id=agent_id, name=agent_name)} && echo "exists"'
+                host, f"test -e {file['path'].format(id=agent_id, name=agent_name)} && echo 'exists'"
             )
-            assert not result, f'This file should not exist in {host} but it was found: ' \
-                               f'{file["path"].format(id=agent_id, name=agent_name)}'
+            assert not result, f"This file should not exist in {host} but it was found: " \
+                               f"{file['path'].format(id=agent_id, name=agent_name)}"
 
-    # Check that agent information does not exist anymore in global.db
+    # Check that agent information is not in the wdb socket
     for host in managers_hosts:
-        for query in db_queries:
-            result = host_manager.run_command(
-                host,
-                f'sqlite3 {join(WAZUH_PATH, "queue", "db", "global.db")} "{query.format(id=agent_id, name=agent_name)}"'
-            )
-            assert not result, f'This db query should have not returned anything in {host}, but it did: ' \
-                               f'{query.format(id=agent_id, name=agent_name)} -> {result}'
+        for query in queries:
+            result = host_manager.run_command(host,
+                                              f"{WAZUH_PATH}/framework/python/bin/python3.9 "
+                                              f"{script_path} '{query.format(id=agent_id)}'")
+            assert not result, f"This db query should have not returned anything in {host}, but it did: {result}"
+
+    host_manager.control_service(host=agent_host, service='wazuh', state='restarted')
+    agent_healthcheck(master_token)