def get_files_status(node_type, get_md5=True):
"""Get all files and metadata inside the directories listed in cluster.json['files'].
Parameters
----------
node_type : str
TODO - To be deprecated
get_md5 : bool
Whether to calculate and save the MD5 hash of the found file.
Returns
-------
final_items : dict
Paths (keys) and metadata (values) of all the files requested in cluster.json['files'].
"""
cluster_items = get_cluster_items()
final_items = {}
for file_path, item in cluster_items['files'].items():
if file_path == "excluded_files" or file_path == "excluded_extensions":
continue
if item['source'] == node_type or item['source'] == 'all':
try:
final_items.update(
walk_dir(file_path, item['recursive'], item['files'],
cluster_items['files']['excluded_files'],
cluster_items['files']['excluded_extensions'],
file_path, get_md5, node_type))
except Exception as e:
logger.warning("Error getting file status: {}.".format(e))
return final_items
def get_files_status(node_type, node_name, get_md5=True):
cluster_items = get_cluster_items()
final_items = {}
for file_path, item in cluster_items['files'].items():
if file_path == "excluded_files" or file_path == "excluded_extensions":
continue
if item['source'] == node_type or item['source'] == 'all':
if item.get("files") and "agent-info.merged" in item["files"]:
agents_to_send, merged_path = \
merge_agent_info(merge_type="agent-info",
node_name=node_name,
time_limit_seconds=cluster_items['sync_options']['get_agentinfo_newer_than']
)
if agents_to_send == 0:
return {}
fullpath = path.dirname(merged_path)
else:
fullpath = file_path
try:
final_items.update(
walk_dir(fullpath, item['recursive'], item['files'],
cluster_items['files']['excluded_files'],
cluster_items['files']['excluded_extensions'],
file_path, get_md5, node_type))
except Exception as e:
logger.warning("Error getting file status: {}.".format(e))
return final_items
def get_files_status(get_md5=True):
"""Get all files and metadata inside the directories listed in cluster.json['files'].
Parameters
----------
get_md5 : bool
Whether to calculate and save the MD5 hash of the found file.
Returns
-------
final_items : dict
Paths (keys) and metadata (values) of all the files requested in cluster.json['files'].
"""
cluster_items = get_cluster_items()
final_items = {}
for file_path, item in cluster_items['files'].items():
if file_path == "excluded_files" or file_path == "excluded_extensions":
continue
try:
final_items.update(
walk_dir(file_path, item['recursive'], item['files'], cluster_items['files']['excluded_files'],
cluster_items['files']['excluded_extensions'], file_path, get_md5))
except Exception as e:
logger.warning(f"Error getting file status: {e}.")
# Save the information collected in the current integration process.
common.cluster_integrity_mtime.set(final_items)
return final_items
def get_cluster_items_worker_intervals():
"""Get worker's time intervals specified in cluster.json file.
Returns
-------
dict
Worker's time intervals specified in cluster.json file.
"""
return get_cluster_items()['intervals']['worker']
def get_cluster_items_communication_intervals():
"""Get communication's time intervals specified in cluster.json file.
Returns
-------
dict
Communication's time intervals specified in cluster.json file.
"""
return get_cluster_items()['intervals']['communication']
def test_get_cluster_items():
"""Verify the cluster files information."""
utils.get_cluster_items.cache_clear()
with patch('os.path.abspath', side_effect=FileNotFoundError):
with pytest.raises(WazuhException, match='.* 3005 .*'):
utils.get_cluster_items()
items = utils.get_cluster_items()
assert items == {'files': {'etc/': {'permissions': 416, 'source': 'master', 'files': ['client.keys'],
'recursive': False, 'restart': False, 'remove_subdirs_if_empty': False,
'extra_valid': False, 'description': 'client keys file database'},
'etc/shared/': {'permissions': 432, 'source': 'master', 'files': ['merged.mg'],
'recursive': True, 'restart': False, 'remove_subdirs_if_empty': True,
'extra_valid': False, 'description': 'shared configuration files'},
'var/multigroups/': {'permissions': 432, 'source': 'master', 'files': ['merged.mg'],
'recursive': True, 'restart': False,
'remove_subdirs_if_empty': True, 'extra_valid': False,
'description': 'shared configuration files'},
'etc/rules/': {'permissions': 432, 'source': 'master', 'files': ['all'],
'recursive': True, 'restart': True, 'remove_subdirs_if_empty': False,
'extra_valid': False, 'description': 'user rules'},
'etc/decoders/': {'permissions': 432, 'source': 'master', 'files': ['all'],
'recursive': True, 'restart': True, 'remove_subdirs_if_empty': False,
'extra_valid': False, 'description': 'user decoders'},
'etc/lists/': {'permissions': 432, 'source': 'master', 'files': ['all'],
'recursive': True, 'restart': True, 'remove_subdirs_if_empty': False,
'extra_valid': False, 'description': 'user CDB lists'},
'queue/agent-groups/': {'permissions': 432, 'source': 'master', 'files': ['all'],
'recursive': True, 'restart': False,
'remove_subdirs_if_empty': False, 'extra_valid': True,
'description': 'agents group configuration'},
'excluded_files': ['ar.conf', 'ossec.conf'],
'excluded_extensions': ['~', '.tmp', '.lock', '.swp']},
'intervals': {'worker': {'sync_integrity': 9, 'sync_agent_info': 10, 'sync_agent_info_ko_retry': 1,
'keep_alive': 60, 'connection_retry': 10,
'max_failed_keepalive_attempts': 2},
'master': {'recalculate_integrity': 8, 'check_worker_lastkeepalive': 60,
'max_allowed_time_without_keepalive': 120},
'communication': {'timeout_cluster_request': 20, 'timeout_api_request': 200,
'timeout_api_exe': 10, 'timeout_receiving_file': 120}},
'sync_options': {'get_agentinfo_newer_than': 1800}, 'distributed_api': {'enabled': True}}
def get_files_status(node_type, get_md5=True):
cluster_items = get_cluster_items()
final_items = {}
for file_path, item in cluster_items['files'].items():
if file_path == "excluded_files" or file_path == "excluded_extensions":
continue
if item['source'] == node_type or item['source'] == 'all':
try:
final_items.update(
walk_dir(file_path, item['recursive'], item['files'],
cluster_items['files']['excluded_files'],
cluster_items['files']['excluded_extensions'],
file_path, get_md5, node_type))
except Exception as e:
logger.warning("Error getting file status: {}.".format(e))
return final_items
debug_mode = 0
# set correct permissions on cluster.log file
if os.path.exists('{0}/logs/cluster.log'.format(common.wazuh_path)):
os.chown('{0}/logs/cluster.log'.format(common.wazuh_path),
common.wazuh_uid(), common.wazuh_gid())
os.chmod('{0}/logs/cluster.log'.format(common.wazuh_path), 0o660)
main_logger = set_logging(foreground_mode=args.foreground,
debug_mode=debug_mode)
cluster_configuration = cluster_utils.read_config(
config_file=args.config_file)
if cluster_configuration['disabled']:
sys.exit(0)
cluster_items = cluster_utils.get_cluster_items()
try:
wazuh.core.cluster.cluster.check_cluster_config(cluster_configuration)
except Exception as e:
main_logger.error(e)
sys.exit(1)
if args.test_config:
sys.exit(0)
cluster_status = wazuh.core.cluster.utils.get_cluster_status()
if cluster_status['running'] == 'yes':
main_logger.error("Cluster is already running.")
sys.exit(1)
# clean
def compare_files(good_files, check_files, node_name):
"""Compare metadata of the master files with metadata of files sent by a worker node.
Compare the integrity information of each file of the master node against those in the worker node (listed in
cluster.json), calculated in get_files_status(). The files are classified in four groups depending on the
information of cluster.json: missing, extra, extra_valid and shared.
Parameters
----------
good_files : dict
Paths (keys) and metadata (values) of the master's files.
check_files : dict
Paths (keys) and metadata (values) of the worker's files.
node_name : str
Name of the worker whose files are being compared.
Returns
-------
files : dict
Paths (keys) and metadata (values) of the files classified into four groups.
count : int
Number of files inside each classification.
"""
def split_on_condition(seq, condition):
"""Split a sequence into two generators based on a condition.
Parameters
----------
seq : set
Set of items to split.
condition : callable
Function base splitting on.
Returns
-------
generator
Items that meet the condition.
generator
Items that do not meet the condition.
"""
l1, l2 = itertools.tee((condition(item), item) for item in seq)
return (i for p, i in l1 if p), (i for p, i in l2 if not p)
# Get 'files' dictionary inside cluster.json.
cluster_items = get_cluster_items()['files']
# Missing files will be the ones that are present in good files (master) but not in the check files (worker).
missing_files = {
key: good_files[key]
for key in good_files.keys() - check_files.keys()
}
# Extra files are the ones present in check files (worker) but not in good files (master) and aren't extra valid.
extra_valid, extra = split_on_condition(
check_files.keys() - good_files.keys(), lambda x: cluster_items[
check_files[x]['cluster_item_key']]['extra_valid'])
extra_files = {key: check_files[key] for key in extra}
extra_valid_files = {key: check_files[key] for key in extra_valid}
# 'all_shared' files are the ones present in both sets but with different MD5 checksum.
all_shared = [
x for x in check_files.keys() & good_files.keys()
if check_files[x]['md5'] != good_files[x]['md5']
]
# 'shared_e_v' are files present in both nodes but need to be merged before sending them to the worker. Only
# 'agent-groups' files fit into this category.
# 'shared' files can be sent as is, without merging.
shared_e_v, shared = split_on_condition(
all_shared, lambda x: cluster_items[check_files[x]['cluster_item_key']]
['extra_valid'])
shared_e_v = list(shared_e_v)
if shared_e_v:
# Merge all shared extra valid files into a single one. Create a tuple (merged_filepath, {metadata_dict}).
shared_merged = [(merge_info(merge_type='agent-groups',
files=shared_e_v,
file_type='-shared',
node_name=node_name,
time_limit_seconds=0)[1], {
'cluster_item_key':
'/queue/agent-groups/',
'merged': True,
'merge-type': 'agent-groups'
})]
# Dict merging all 'shared' filepaths (keys) and the merged_filepath (key) created above.
shared_files = dict(
itertools.chain(shared_merged,
((key, good_files[key]) for key in shared)))
else:
shared_files = {key: good_files[key] for key in shared}
files = {
'missing': missing_files,
'extra': extra_files,
'shared': shared_files,
'extra_valid': extra_valid_files
}
count = {
'missing': len(missing_files),
'extra': len(extra_files),
'extra_valid': len(extra_valid_files),
'shared': len(all_shared)
}
return files, count
def get_cluster_items_worker_intervals():
return get_cluster_items()['intervals']['worker']
def get_cluster_items_communication_intervals():
return get_cluster_items()['intervals']['communication']
def get_cluster_items_master_intervals():
return get_cluster_items()['intervals']['master']
def compare_files(good_files, check_files, node_name):
def split_on_condition(seq, condition):
"""
Splits a sequence into two generators based on a conditon
:param seq: sequence to split
:param condition: function base splitting on
:return: two generators
"""
l1, l2 = itertools.tee((condition(item), item) for item in seq)
return (i for p, i in l1 if p), (i for p, i in l2 if not p)
cluster_items = get_cluster_items()['files']
# missing files will be the ones that are present in good files but not in the check files
missing_files = {
key: good_files[key]
for key in good_files.keys() - check_files.keys()
}
# extra files are the ones present in check files but not in good files and aren't extra valid
extra_valid, extra = split_on_condition(
check_files.keys() - good_files.keys(), lambda x: cluster_items[
check_files[x]['cluster_item_key']]['extra_valid'])
extra_files = {key: check_files[key] for key in extra}
extra_valid_files = {key: check_files[key] for key in extra_valid}
# shared files are the ones present in both sets.
all_shared = [
x for x in check_files.keys() & good_files.keys()
if check_files[x]['md5'] != good_files[x]['md5']
]
shared_e_v, shared = split_on_condition(
all_shared, lambda x: cluster_items[check_files[x]['cluster_item_key']]
['extra_valid'])
shared_e_v = list(shared_e_v)
if shared_e_v:
# merge all shared extra valid files into a single one.
# To Do: if more extra valid files types are included, compute their merge type and remove hardcoded
# agent-groups
shared_merged = [(merge_agent_info(merge_type='agent-groups',
files=shared_e_v,
file_type='-shared',
node_name=node_name,
time_limit_seconds=0)[1], {
'cluster_item_key':
'/queue/agent-groups/',
'merged': True,
'merge-type': 'agent-groups'
})]
shared_files = dict(
itertools.chain(shared_merged,
((key, good_files[key]) for key in shared)))
else:
shared_files = {key: good_files[key] for key in shared}
files = {
'missing': missing_files,
'extra': extra_files,
'shared': shared_files,
'extra_valid': extra_valid_files
}
count = {
'missing': len(missing_files),
'extra': len(extra_files),
'extra_valid': len(extra_valid_files),
'shared': len(all_shared)
}
return files, count
