def wait(self, promise, timeout=None):
e = Event()
def on_resolve_or_reject(_):
e.set()
promise._then(on_resolve_or_reject, on_resolve_or_reject)
# We can't use the timeout in Asyncio event
e.wait()
async def event_wait(event: asyncio.Event, timeout=None):
'''
Wait on an an asyncio event with an optional timeout
Returns:
true if the event got set, None if timed out
'''
if timeout is None:
await event.wait()
return True
try:
await asyncio.wait_for(event.wait(), timeout)
except asyncio.TimeoutError:
return False
return True
class Actor(object):
"""
Main actor model.
Args:
inbox (GeneratorQueue): Inbox to consume from.
outbox (GeneratorQueue): Outbox to publish to.
loop (GeneratorQueue): Event loop.
"""
running = False
_force_stop = False
def __init__(self, inbox, outbox, loop=None):
self.inbox = inbox
self.outbox = outbox
if not loop:
loop = get_event_loop()
self._loop = loop
self._pause_lock = Lock(loop=self._loop)
self._stop_event = Event(loop=self._loop)
self._test = None
self.__testy = None
self.on_init()
@property
def paused(self):
"""Indicate if actor is paused."""
return self._pause_lock.locked()
async def start(self):
"""Main public entry point to start the actor."""
await self.initialize()
await self._start()
async def initialize(self):
"""Initialize the actor before starting."""
await self.on_start()
if self._force_stop:
return
if not self.running:
self.running = True
async def _start(self):
"""Run the event loop and force the on_stop event."""
try:
await self._run()
finally:
await self.on_stop()
async def resume(self):
"""Resume the actor."""
await self.on_resume()
self._pause_lock.release()
async def pause(self):
"""Pause the actor."""
await self._pause_lock.acquire()
await self.on_pause()
async def _block_if_paused(self):
"""Block on the pause lock."""
if self.paused:
await self._pause_lock.acquire()
await self._pause_lock.release()
async def _run(self):
"""Main event loop."""
while self.running:
await self._block_if_paused()
await self._process()
async def publish(self, data):
"""Push data to the outbox."""
await self.outbox.put(data)
async def stop(self):
"""Stop the actor."""
self.inbox = None
self.outbox = None
self.running = False
self._force_stop = True
self._stop_event.set()
try:
self._pause_lock.release()
except RuntimeError:
pass
async def _process(self):
"""Process incoming messages."""
if not self.inbox:
return
pending = {self.inbox.get(), self._stop_event.wait()}
result = await get_first_completed(pending, self._loop)
if self.running:
await self.on_message(result)
async def on_message(self, data):
"""Called when the actor receives a message."""
raise NotImplementedError
def on_init(self):
"""Called after the actor class is instantiated."""
pass
async def on_start(self):
"""Called before the actor starts ingesting the inbox."""
pass
async def on_stop(self):
"""Called after actor dies."""
pass
async def on_pause(self):
"""Called before the actor is paused."""
pass
async def on_resume(self):
"""Called before the actor is resumed."""
pass
class SubscribeListener(SubscribeCallback):
def __init__(self):
self.connected = False
self.connected_event = Event()
self.disconnected_event = Event()
self.presence_queue = Queue()
self.message_queue = Queue()
self.error_queue = Queue()
def status(self, pubnub, status):
if utils.is_subscribed_event(status) and not self.connected_event.is_set():
self.connected_event.set()
elif utils.is_unsubscribed_event(status) and not self.disconnected_event.is_set():
self.disconnected_event.set()
elif status.is_error():
self.error_queue.put_nowait(status.error_data.exception)
def message(self, pubnub, message):
self.message_queue.put_nowait(message)
def presence(self, pubnub, presence):
self.presence_queue.put_nowait(presence)
@asyncio.coroutine
def _wait_for(self, coro):
scc_task = asyncio.ensure_future(coro)
err_task = asyncio.ensure_future(self.error_queue.get())
yield from asyncio.wait([
scc_task,
err_task
], return_when=asyncio.FIRST_COMPLETED)
if err_task.done() and not scc_task.done():
if not scc_task.cancelled():
scc_task.cancel()
raise err_task.result()
else:
if not err_task.cancelled():
err_task.cancel()
return scc_task.result()
@asyncio.coroutine
def wait_for_connect(self):
if not self.connected_event.is_set():
yield from self._wait_for(self.connected_event.wait())
else:
raise Exception("instance is already connected")
@asyncio.coroutine
def wait_for_disconnect(self):
if not self.disconnected_event.is_set():
yield from self._wait_for(self.disconnected_event.wait())
else:
raise Exception("instance is already disconnected")
@asyncio.coroutine
def wait_for_message_on(self, *channel_names):
channel_names = list(channel_names)
while True:
try:
env = yield from self._wait_for(self.message_queue.get())
if env.channel in channel_names:
return env
else:
continue
finally:
self.message_queue.task_done()
@asyncio.coroutine
def wait_for_presence_on(self, *channel_names):
channel_names = list(channel_names)
while True:
try:
env = yield from self._wait_for(self.presence_queue.get())
if env.channel in channel_names:
return env
else:
continue
finally:
self.presence_queue.task_done()
class SessionBase(asyncio.Protocol):
"""Base class of networking sessions.
There is no client / server distinction other than who initiated
the connection.
To initiate a connection to a remote server pass host, port and
proxy to the constructor, and then call create_connection(). Each
successful call should have a corresponding call to close().
Alternatively if used in a with statement, the connection is made
on entry to the block, and closed on exit from the block.
"""
max_errors = 10
def __init__(self, *, framer=None, loop=None):
self.framer = framer or self.default_framer()
self.loop = loop or asyncio.get_event_loop()
self.logger = logging.getLogger(self.__class__.__name__)
self.transport = None
# Set when a connection is made
self._address = None
self._proxy_address = None
# For logger.debug messages
self.verbosity = 0
# Cleared when the send socket is full
self._can_send = Event()
self._can_send.set()
self._pm_task = None
self._task_group = TaskGroup(self.loop)
# Force-close a connection if a send doesn't succeed in this time
self.max_send_delay = 60
# Statistics. The RPC object also keeps its own statistics.
self.start_time = time.perf_counter()
self.errors = 0
self.send_count = 0
self.send_size = 0
self.last_send = self.start_time
self.recv_count = 0
self.recv_size = 0
self.last_recv = self.start_time
self.last_packet_received = self.start_time
async def _limited_wait(self, secs):
try:
await asyncio.wait_for(self._can_send.wait(), secs)
except asyncio.TimeoutError:
self.abort()
raise asyncio.TimeoutError(f'task timed out after {secs}s')
async def _send_message(self, message):
if not self._can_send.is_set():
await self._limited_wait(self.max_send_delay)
if not self.is_closing():
framed_message = self.framer.frame(message)
self.send_size += len(framed_message)
self.send_count += 1
self.last_send = time.perf_counter()
if self.verbosity >= 4:
self.logger.debug(f'Sending framed message {framed_message}')
self.transport.write(framed_message)
def _bump_errors(self):
self.errors += 1
if self.errors >= self.max_errors:
# Don't await self.close() because that is self-cancelling
self._close()
def _close(self):
if self.transport:
self.transport.close()
# asyncio framework
def data_received(self, framed_message):
"""Called by asyncio when a message comes in."""
self.last_packet_received = time.perf_counter()
if self.verbosity >= 4:
self.logger.debug(f'Received framed message {framed_message}')
self.recv_size += len(framed_message)
self.framer.received_bytes(framed_message)
def pause_writing(self):
"""Transport calls when the send buffer is full."""
if not self.is_closing():
self._can_send.clear()
self.transport.pause_reading()
def resume_writing(self):
"""Transport calls when the send buffer has room."""
if not self._can_send.is_set():
self._can_send.set()
self.transport.resume_reading()
def connection_made(self, transport):
"""Called by asyncio when a connection is established.
Derived classes overriding this method must call this first."""
self.transport = transport
# This would throw if called on a closed SSL transport. Fixed
# in asyncio in Python 3.6.1 and 3.5.4
peer_address = transport.get_extra_info('peername')
# If the Socks proxy was used then _address is already set to
# the remote address
if self._address:
self._proxy_address = peer_address
else:
self._address = peer_address
self._pm_task = self.loop.create_task(self._receive_messages())
def connection_lost(self, exc):
"""Called by asyncio when the connection closes.
Tear down things done in connection_made."""
self._address = None
self.transport = None
self._task_group.cancel()
if self._pm_task:
self._pm_task.cancel()
# Release waiting tasks
self._can_send.set()
# External API
def default_framer(self):
"""Return a default framer."""
raise NotImplementedError
def peer_address(self):
"""Returns the peer's address (Python networking address), or None if
no connection or an error.
This is the result of socket.getpeername() when the connection
was made.
"""
return self._address
def peer_address_str(self):
"""Returns the peer's IP address and port as a human-readable
string."""
if not self._address:
return 'unknown'
ip_addr_str, port = self._address[:2]
if ':' in ip_addr_str:
return f'[{ip_addr_str}]:{port}'
else:
return f'{ip_addr_str}:{port}'
def is_closing(self):
"""Return True if the connection is closing."""
return not self.transport or self.transport.is_closing()
def abort(self):
"""Forcefully close the connection."""
if self.transport:
self.transport.abort()
# TODO: replace with synchronous_close
async def close(self, *, force_after=30):
"""Close the connection and return when closed."""
self._close()
if self._pm_task:
with suppress(CancelledError):
await asyncio.wait([self._pm_task], timeout=force_after)
self.abort()
await self._pm_task
def synchronous_close(self):
self._close()
if self._pm_task and not self._pm_task.done():
self._pm_task.cancel()
class AsyncConnectionPool:
"""Object manages asynchronous connections.
:param int size: size (number of connection) of the pool.
:param float queue_timeout: time out when client is waiting connection
from pool
:param loop: event loop, if not passed then default will be used
:param config: MySql connection config see
`doc. <http://dev.mysql.com/doc/connector-python/en/connector-python-connectargs.html>`_
:raise ValueError: if the `size` is inappropriate
"""
def __init__(self, size=1, queue_timeout=15.0, *, loop=None, **config):
assert size > 0, 'DBPool.size must be greater than 0'
if size < 1:
raise ValueError('DBPool.size is less than 1, '
'connections won"t be established')
self._pool = set()
self._busy_items = set()
self._size = size
self._pending_futures = deque()
self._queue_timeout = queue_timeout
self._loop = loop or asyncio.get_event_loop()
self.config = config
self._shutdown_event = Event(loop=self._loop)
self._shutdown_event.set()
@property
def queue_timeout(self):
"""Number of seconds to wait a connection from the pool,
before TimeoutError occurred
:rtype: float
"""
return self._queue_timeout
@queue_timeout.setter
def queue_timeout(self, value):
"""Sets a timeout for :attr:`queue_timeout`
:param float value: number of seconds
"""
if not isinstance(value, (float, int)):
raise ValueError('Float or integer type expected')
self._queue_timeout = value
@property
def size(self):
"""Size of pool
:rtype: int
"""
return self._size
def __len__(self):
"""Number of allocated pool's slots
:rtype: int
"""
return len(self._pool)
@property
def free_count(self):
"""Number of free pool's slots
:rtype: int
"""
return self.size - len(self._busy_items)
@asyncio.coroutine
def get(self):
"""Coroutine. Returns an opened connection from pool.
If coroutine invoked when all connections have been issued, then
caller will blocked until some connection will be released.
Also, the class provides context manager for getting connection
and automatically freeing it. Example:
>>> with (yield from pool) as cnx:
>>> ...
:rtype: AsyncMySQLConnection
:raise: concurrent.futures.TimeoutError()
"""
cnx = None
yield from self._shutdown_event.wait()
for free_client in self._pool - self._busy_items:
cnx = free_client
self._busy_items.add(cnx)
break
else:
if len(self) < self.size:
cnx = AsyncMySQLConnection(loop=self._loop)
self._pool.add(cnx)
self._busy_items.add(cnx)
try:
yield from cnx.connect(**self.config)
except:
self._pool.remove(cnx)
self._busy_items.remove(cnx)
raise
if not cnx:
queue_future = Future(loop=self._loop)
self._pending_futures.append(queue_future)
try:
cnx = yield from asyncio.wait_for(queue_future,
self.queue_timeout,
loop=self._loop)
self._busy_items.add(cnx)
except TimeoutError:
raise TimeoutError('Database pool is busy')
finally:
try:
self._pending_futures.remove(queue_future)
except ValueError:
pass
return cnx
def release(self, connection):
"""Frees connection. After that the connection can be issued
by :func:`get`.
:param AsyncMySQLConnection connection: a connection received
from :func:`get`
"""
if len(self._pending_futures):
f = self._pending_futures.popleft()
f.set_result(connection)
else:
self._busy_items.remove(connection)
@asyncio.coroutine
def shutdown(self):
"""Coroutine. Closes all connections and purge queue of a waiting
for connection.
"""
self._shutdown_event.clear()
try:
for cnx in self._pool:
yield from cnx.disconnect()
for f in self._pending_futures:
f.cancel()
self._pending_futures.clear()
self._pool = set()
self._busy_items = set()
finally:
self._shutdown_event.set()
def __enter__(self):
raise RuntimeError(
'"yield from" should be used as context manager expression')
def __exit__(self, *args):
# This must exist because __enter__ exists, even though that
# always raises; that's how the with-statement works.
pass
@asyncio.coroutine
def __iter__(self):
cnx = yield from self.get()
return ContextManager(self, cnx)
async def process_files_from_master(self, name: str, file_received: asyncio.Event):
"""Perform relevant actions for each file according to its status.
Process integrity files coming from the master. It updates necessary information and sends the master
any required extra_valid files.
Parameters
----------
name : str
Task ID that was waiting for the file to be received.
file_received : asyncio.Event
Asyncio event that is unlocked once the file has been received.
"""
logger = self.task_loggers['Integrity sync']
try:
await asyncio.wait_for(file_received.wait(),
timeout=self.cluster_items['intervals']['communication']['timeout_receiving_file'])
except Exception:
await self.send_request(
command=b'syn_i_w_m_r',
data=b'None ' + json.dumps(timeout_exc := WazuhClusterError(
3039, extra_message=f'Integrity sync at {self.name}'), cls=c_common.WazuhJSONEncoder).encode())
raise timeout_exc
if isinstance(self.sync_tasks[name].filename, Exception):
exc_info = json.dumps(exception.WazuhClusterError(
1000, extra_message=str(self.sync_tasks[name].filename)), cls=c_common.WazuhJSONEncoder)
await self.send_request(command=b'syn_i_w_m_r', data=b'None ' + exc_info.encode())
raise self.sync_tasks[name].filename
zip_path = ""
# Path of the zip containing a JSON with metadata and files to be updated in this worker node.
received_filename = self.sync_tasks[name].filename
try:
self.integrity_sync_status['date_start'] = datetime.utcnow().timestamp()
logger.info("Starting.")
"""
- zip_path contains the path of the unzipped directory
- ko_files contains a Dict with this structure:
{'missing': {'<file_path>': {<MD5, merged, merged_name, etc>}, ...},
'shared': {...}, 'extra': {...}, 'extra_valid': {...}}
"""
ko_files, zip_path = await cluster.run_in_pool(self.loop, self.manager.task_pool, cluster.decompress_files,
received_filename)
logger.info("Files to create: {} | Files to update: {} | Files to delete: {} | Files to send: {}".format(
len(ko_files['missing']), len(ko_files['shared']), len(ko_files['extra']), len(ko_files['extra_valid']))
)
if ko_files['shared'] or ko_files['missing'] or ko_files['extra']:
# Update or remove files in this worker node according to their status (missing, extra or shared).
logger.debug("Worker does not meet integrity checks. Actions required.")
logger.debug("Updating local files: Start.")
await cluster.run_in_pool(self.loop, self.manager.task_pool, self.update_master_files_in_worker,
ko_files, zip_path, self.cluster_items, self.task_loggers['Integrity sync'])
logger.debug("Updating local files: End.")
# Send extra valid files to the master.
if ko_files['extra_valid']:
logger.debug("Master requires some worker files.")
asyncio.create_task(self.sync_extra_valid(ko_files['extra_valid']))
else:
logger.info(
f"Finished in {datetime.utcnow().timestamp() - self.integrity_sync_status['date_start']:.3f}s.")
except exception.WazuhException as e:
logger.error(f"Error synchronizing extra valid files: {e}")
await self.send_request(command=b'syn_i_w_m_r',
data=b'None ' + json.dumps(e, cls=c_common.WazuhJSONEncoder).encode())
except Exception as e:
logger.error(f"Error synchronizing extra valid files: {e}")
exc_info = json.dumps(exception.WazuhClusterError(1000, extra_message=str(e)),
cls=c_common.WazuhJSONEncoder)
await self.send_request(command=b'syn_i_w_m_r', data=b'None ' + exc_info.encode())
finally:
zip_path and shutil.rmtree(zip_path)
class SubscribeListener(SubscribeCallback):
def __init__(self):
self.connected = False
self.connected_event = Event()
self.disconnected_event = Event()
self.presence_queue = Queue()
self.message_queue = Queue()
self.error_queue = Queue()
def status(self, pubnub, status):
if utils.is_subscribed_event(
status) and not self.connected_event.is_set():
self.connected_event.set()
elif utils.is_unsubscribed_event(
status) and not self.disconnected_event.is_set():
self.disconnected_event.set()
elif status.is_error():
self.error_queue.put_nowait(status.error_data.exception)
def message(self, pubnub, message):
self.message_queue.put_nowait(message)
def presence(self, pubnub, presence):
self.presence_queue.put_nowait(presence)
async def _wait_for(self, coro):
scc_task = asyncio.ensure_future(coro)
err_task = asyncio.ensure_future(self.error_queue.get())
await asyncio.wait([scc_task, err_task],
return_when=asyncio.FIRST_COMPLETED)
if err_task.done() and not scc_task.done():
if not scc_task.cancelled():
scc_task.cancel()
raise err_task.result()
else:
if not err_task.cancelled():
err_task.cancel()
return scc_task.result()
async def wait_for_connect(self):
if not self.connected_event.is_set():
await self._wait_for(self.connected_event.wait())
else:
raise Exception("instance is already connected")
async def wait_for_disconnect(self):
if not self.disconnected_event.is_set():
await self._wait_for(self.disconnected_event.wait())
else:
raise Exception("instance is already disconnected")
async def wait_for_message_on(self, *channel_names):
channel_names = list(channel_names)
while True:
try:
env = await self._wait_for(self.message_queue.get())
if env.channel in channel_names:
return env
else:
continue
finally:
self.message_queue.task_done()
async def wait_for_presence_on(self, *channel_names):
channel_names = list(channel_names)
while True:
try:
env = await self._wait_for(self.presence_queue.get())
if env.channel in channel_names:
return env
else:
continue
finally:
self.presence_queue.task_done()
async def connect(self, addr, port, timeout=3):
self.logger.debug('hxsocks2 send connect request')
async with self._lock:
if self.connection_lost:
self._manager.remove(self)
raise ConnectionResetError(0, 'hxs connection lost')
if not self.connected:
try:
await self.get_key(timeout)
except asyncio.CancelledError:
raise
except Exception as err:
self.logger.error('%s get_key %r', self.name, err)
# self.logger.error(traceback.format_exc())
try:
self.remote_writer.close()
except (OSError, AttributeError):
pass
raise ConnectionResetError(0, 'hxs get_key failed.')
# send connect request
payload = b''.join([
chr(len(addr)).encode('latin1'),
addr.encode(),
struct.pack('>H', port),
b'\x00' * random.randint(64, 255),
])
stream_id = self._next_stream_id
self._next_stream_id += 1
if self._next_stream_id > MAX_STREAM_ID:
self.logger.error('MAX_STREAM_ID reached')
self._manager.remove(self)
await self.send_frame(1, 0, stream_id, payload)
# wait for server response
event = Event()
self._client_status[stream_id] = event
# await event.wait()
fut = event.wait()
try:
await asyncio.wait_for(fut, timeout=timeout)
except asyncio.TimeoutError:
self.logger.error('no response from %s, timeout=%.3f', self.name,
timeout)
del self._client_status[stream_id]
self.print_status()
await self.send_ping()
raise
del self._client_status[stream_id]
if self._stream_status[stream_id] == OPEN:
socketpair_a, socketpair_b = socket.socketpair()
if sys.platform == 'win32':
socketpair_a.setsockopt(socket.SOL_SOCKET, socket.SO_SNDBUF,
65536)
socketpair_b.setsockopt(socket.SOL_SOCKET, socket.SO_SNDBUF,
65536)
reader, writer = await asyncio.open_connection(sock=socketpair_b)
writer.transport.set_write_buffer_limits(0, 0)
self._client_writer[stream_id] = writer
self._last_active[stream_id] = time.monotonic()
# start forwarding
asyncio.ensure_future(self.read_from_client(stream_id, reader))
return socketpair_a
raise ConnectionResetError(
0, 'remote connect to %s:%d failed.' % (addr, port))
class Channel(object):
"""
A Channel is a closable queue. A Channel is considered "finished" when
it is closed and drained (unlike a queue which is "finished" when the queue
is empty)
"""
def __init__(self, maxsize=0, *, loop=None):
if loop is None:
self._loop = get_event_loop()
else:
self._loop = loop
if not isinstance(maxsize, int) or maxsize < 0:
raise TypeError("maxsize must be an integer >= 0 (default is 0)")
self._maxsize = maxsize
# Futures.
self._getters = deque()
self._putters = deque()
# "finished" means channel is closed and drained
self._finished = Event(loop=self._loop)
self._close = Event(loop=self._loop)
self._init()
def _init(self):
self._queue = deque()
def _get(self):
return self._queue.popleft()
def _put(self, item):
self._queue.append(item)
def _wakeup_next(self, waiters):
# Wake up the next waiter (if any) that isn't cancelled.
while waiters:
waiter = waiters.popleft()
if not waiter.done():
waiter.set_result(None)
break
def __repr__(self):
return '<{} at {:#x} maxsize={!r} qsize={!r}>'.format(
type(self).__name__, id(self), self._maxsize, self.qsize())
def __str__(self):
return '<{} maxsize={!r} qsize={!r}>'.format(
type(self).__name__, self._maxsize, self.qsize())
def qsize(self):
"""Number of items in the channel buffer."""
return len(self._queue)
@property
def maxsize(self):
"""Number of items allowed in the channel buffer."""
return self._maxsize
def empty(self):
"""Return True if the channel is empty, False otherwise."""
return not self._queue
def full(self):
"""Return True if there are maxsize items in the channel.
Note: if the Channel was initialized with maxsize=0 (the default),
then full() is never True.
"""
if self._maxsize <= 0:
return False
else:
return self.qsize() >= self._maxsize
@coroutine
def put(self, item):
"""Put an item into the channel.
If the channel is full, wait until a free
slot is available before adding item.
If the channel is closed or closing, raise ChannelClosed.
This method is a coroutine.
"""
while self.full() and not self._close.is_set():
putter = Future(loop=self._loop)
self._putters.append(putter)
try:
yield from putter
except ChannelClosed:
raise
except:
putter.cancel() # Just in case putter is not done yet.
if not self.full() and not putter.cancelled():
# We were woken up by get_nowait(), but can't take
# the call. Wake up the next in line.
self._wakeup_next(self._putters)
raise
return self.put_nowait(item)
def put_nowait(self, item):
"""Put an item into the channel without blocking.
If no free slot is immediately available, raise ChannelFull.
"""
if self.full():
raise ChannelFull
if self._close.is_set():
raise ChannelClosed
self._put(item)
self._wakeup_next(self._getters)
@coroutine
def get(self):
"""Remove and return an item from the channel.
If channel is empty, wait until an item is available.
This method is a coroutine.
"""
while self.empty() and not self._close.is_set():
getter = Future(loop=self._loop)
self._getters.append(getter)
try:
yield from getter
except ChannelClosed:
raise
except:
getter.cancel() # Just in case getter is not done yet.
if not self.empty() and not getter.cancelled():
# We were woken up by put_nowait(), but can't take
# the call. Wake up the next in line.
self._wakeup_next(self._getters)
raise
return self.get_nowait()
def get_nowait(self):
"""Remove and return an item from the channel.
Return an item if one is immediately available, else raise ChannelEmpty.
"""
if self.empty():
if self._close.is_set():
raise ChannelClosed
else:
raise ChannelEmpty
item = self._get()
if self.empty() and self._close.is_set():
# if empty _after_ we retrieved an item AND marked for closing,
# set the finished flag
self._finished.set()
self._wakeup_next(self._putters)
return item
@coroutine
def join(self):
"""Block until channel is closed and channel is drained
"""
yield from self._finished.wait()
def close(self):
"""Marks the channel is closed and throw a ChannelClosed in all pending putters"""
self._close.set()
# cancel putters
for putter in self._putters:
putter.set_exception(ChannelClosed())
# cancel getters that can't ever return (as no more items can be added)
while len(self._getters) > self.qsize():
getter = self._getters.pop()
getter.set_exception(ChannelClosed())
if self.empty():
# already empty, mark as finished
self._finished.set()
def closed(self):
"""Returns True if the Channel is marked as closed"""
return self._close.is_set()
@coroutine
def __aiter__(self): # pragma: no cover
"""Returns an async iterator (self)"""
return self
@coroutine
def __anext__(self): # pragma: no cover
try:
data = yield from self.get()
except ChannelClosed:
raise StopAsyncIteration
else:
return data
def __iter__(self):
return iter(self._queue)
async def sync_integrity(self, task_id: str, received_file: asyncio.Event):
"""Perform the integrity synchronization process by comparing local and received files.
It waits until the worker sends its integrity metadata. Once received, they are unzipped.
The information inside the unzipped files_metadata.json file (integrity metadata) is compared with the
local one (updated every self.cluster_items['intervals']['master']['recalculate_integrity'] seconds).
All files that are different (new, deleted, with a different MD5, etc) are classified into four groups:
shared, missing, extra and extra_valid.
Finally, a zip containing this classification (files_metadata.json) and the files that are missing
or that must be updated are sent to the worker.
Parameters
----------
task_id : str
ID of the asyncio task in charge of doing the sync process.
received_file : asyncio.Event
Asyncio event that is holding a lock while the files are not received.
Returns
-------
bytes
Result.
bytes
Response message.
"""
logger = self.task_loggers['Integrity check']
date_start_master = datetime.now()
logger.debug("Waiting to receive zip file from worker.")
await asyncio.wait_for(received_file.wait(),
timeout=self.cluster_items['intervals']['communication']['timeout_receiving_file'])
# Full path where the zip sent by the worker is located.
received_filename = self.sync_tasks[task_id].filename
if isinstance(received_filename, Exception):
raise received_filename
logger.debug(f"Received file from worker: '{received_filename}'")
# Path to metadata file (files_metadata.json) and to zipdir (directory with decompressed files).
files_metadata, decompressed_files_path = await wazuh.core.cluster.cluster.decompress_files(received_filename)
# There are no files inside decompressed_files_path, only files_metadata.json which has already been loaded.
shutil.rmtree(decompressed_files_path)
logger.info(f"Starting. Received metadata of {len(files_metadata)} files.")
# Classify files in shared, missing, extra and extra valid.
worker_files_ko, counts = wazuh.core.cluster.cluster.compare_files(self.server.integrity_control,
files_metadata, self.name)
self.integrity_check_status.update({'date_start_master': date_start_master, 'date_end_master': datetime.now()})
total_time = (self.integrity_check_status['date_end_master'] - date_start_master).total_seconds()
# Get the total number of files that require some change.
if not functools.reduce(operator.add, map(len, worker_files_ko.values())):
logger.info(f"Finished in {total_time:.3f}s. Sync not required.")
result = await self.send_request(command=b'syn_m_c_ok', data=b'')
else:
logger.info(f"Finished in {total_time:.3f}s. Sync required.")
logger = self.task_loggers['Integrity sync']
logger.info("Starting.")
self.integrity_sync_status.update({'tmp_date_start_master': datetime.now(), 'total_files': counts,
'total_extra_valid': 0})
logger.info("Files to create in worker: {} | Files to update in worker: {} | Files to delete in worker: {} "
"| Files to receive: {}".format(len(worker_files_ko['missing']), len(worker_files_ko['shared']),
len(worker_files_ko['extra']), len(worker_files_ko['extra_valid']))
)
# Compress data: master files (only KO shared and missing).
logger.debug("Compressing files to be synced in worker.")
master_files_paths = worker_files_ko['shared'].keys() | worker_files_ko['missing'].keys()
compressed_data = wazuh.core.cluster.cluster.compress_files(self.name, master_files_paths, worker_files_ko)
logger.debug("Zip with files to be synced sent to worker.")
try:
# Start the synchronization process with the worker and get a taskID.
task_id = await self.send_request(command=b'syn_m_c', data=b'')
if isinstance(task_id, Exception) or task_id.startswith(b'Error'):
exc_info = task_id if isinstance(task_id, Exception) else \
exception.WazuhClusterError(code=3016, extra_message=str(task_id))
task_id = b'None'
raise exc_info
# Send zip file to the worker into chunks.
await self.send_file(compressed_data)
# Finish the synchronization process and notify where the file corresponding to the taskID is located.
result = await self.send_request(command=b'syn_m_c_e',
data=task_id + b' ' + os.path.relpath(
compressed_data, common.wazuh_path).encode())
if isinstance(result, Exception):
raise result
elif result.startswith(b'Error'):
raise exception.WazuhClusterError(3016, extra_message=result.decode())
except exception.WazuhException as e:
# Notify error to worker and delete its received file.
self.logger.error(f"Error sending files information: {e}")
result = await self.send_request(command=b'syn_m_c_r', data=task_id + b' ' +
json.dumps(e, cls=c_common.WazuhJSONEncoder).encode())
except Exception as e:
# Notify error to worker and delete its received file.
self.logger.error(f"Error sending files information: {e}")
exc_info = json.dumps(exception.WazuhClusterError(code=1000, extra_message=str(e)),
cls=c_common.WazuhJSONEncoder).encode()
result = await self.send_request(command=b'syn_m_c_r', data=task_id + b' ' + exc_info)
finally:
# Remove local file.
os.unlink(compressed_data)
logger.debug("Finished sending files to worker.")
# Log 'Finished in' message only if there are no extra_valid files to sync.
if not worker_files_ko['extra_valid']:
self.integrity_sync_status['date_start_master'] = self.integrity_sync_status['tmp_date_start_master']
self.integrity_sync_status['date_end_master'] = datetime.now()
logger.info("Finished in {:.3f}s.".format((self.integrity_sync_status['date_end_master'] -
self.integrity_sync_status['date_start_master'])
.total_seconds()))
self.sync_integrity_free = True
return result
class MapAsyncIterator:
"""Map an AsyncIterable over a callback function.
Given an AsyncIterable and a callback function, return an AsyncIterator which
produces values mapped via calling the callback function.
When the resulting AsyncIterator is closed, the underlying AsyncIterable will also
be closed.
"""
def __init__(
self,
iterable: AsyncIterable,
callback: Callable,
reject_callback: Optional[Callable] = None,
) -> None:
self.iterator = iterable.__aiter__()
self.callback = callback
self.reject_callback = reject_callback
self._close_event = Event()
def __aiter__(self) -> "MapAsyncIterator":
"""Get the iterator object."""
return self
async def __anext__(self) -> Any:
"""Get the next value of the iterator."""
if self.is_closed:
if not isasyncgen(self.iterator):
raise StopAsyncIteration
value = await self.iterator.__anext__()
result = self.callback(value)
else:
aclose = ensure_future(self._close_event.wait())
anext = ensure_future(self.iterator.__anext__())
try:
pending: Set[Task] = (await
wait([aclose, anext],
return_when=FIRST_COMPLETED))[1]
except CancelledError:
# cancel underlying tasks and close
aclose.cancel()
anext.cancel()
await self.aclose()
raise # re-raise the cancellation
for task in pending:
task.cancel()
if aclose.done():
raise StopAsyncIteration
error = anext.exception()
if error:
if not self.reject_callback or isinstance(
error, (StopAsyncIteration, GeneratorExit)):
raise error
result = self.reject_callback(error)
else:
value = anext.result()
result = self.callback(value)
return await result if isawaitable(result) else result
async def athrow(
self,
type_: Union[BaseException, Type[BaseException]],
value: Optional[BaseException] = None,
traceback: Optional[TracebackType] = None,
) -> None:
"""Throw an exception into the asynchronous iterator."""
if not self.is_closed:
athrow = getattr(self.iterator, "athrow", None)
if athrow:
await athrow(type_, value, traceback)
else:
await self.aclose()
if value is None:
if traceback is None:
raise type_
value = (type_ if isinstance(value, BaseException) else
cast(Type[BaseException], type_)())
if traceback is not None:
value = value.with_traceback(traceback)
raise value
async def aclose(self) -> None:
"""Close the iterator."""
if not self.is_closed:
aclose = getattr(self.iterator, "aclose", None)
if aclose:
try:
await aclose()
except RuntimeError:
pass
self.is_closed = True
@property
def is_closed(self) -> bool:
"""Check whether the iterator is closed."""
return self._close_event.is_set()
@is_closed.setter
def is_closed(self, value: bool) -> None:
"""Mark the iterator as closed."""
if value:
self._close_event.set()
else:
self._close_event.clear()
class PubSubInstance:
__slots__ = '_pubsub', '_encoder', '_decoder', '_closed', '_messages', '_event'
def __init__(self, pubsub, encoder, decoder):
self._pubsub = pubsub
self._encoder = encoder or utf8_encode
self._decoder = decoder
self._closed = False
self._messages = deque()
self._event = Event()
async def __aenter__(self):
if self._closed:
raise RedisError('Pub/sub instance closed')
return self
async def __aexit__(self, exc_type, exc_value, traceback):
await self.aclose()
async def aclose(self):
if not self._closed:
self._closed = True
try:
await self._pubsub.unregister(self)
except Exception:
pass
self._messages = None
self._decoder = None
self._encoder = None
self._pubsub = None
self._event = None
async def add(self, channels=None, patterns=None):
await self._cmd(self._pubsub.register, channels, patterns)
# TODO (question) should we removed the self._messages that are not related to this channels and patterns (left overs)?
async def remove(self, channels=None, patterns=None):
await self._cmd(self._pubsub.unregister, channels, patterns)
async def message(self, timeout=None):
if self._closed:
raise RedisError('Pub/sub instance closed')
msg = self._get_message()
if msg is not None:
return msg
self._event.clear()
# We check connection here to notify the end user if there is an connection error...
await self._pubsub.check_connection(self)
if timeout is None:
await self._event.wait()
else:
try:
await wait_for(self._event.wait(), timeout)
except AsyncIOTimeoutError:
pass
return self._get_message()
async def ping(self, message=None):
if self._closed:
raise RedisError('Pub/sub instance closed')
await self._pubsub.ping(message)
async def _cmd(self, cmd, channels, patterns):
if self._closed:
raise RedisError('Pub/sub instance closed')
if channels:
if isinstance(channels, (str, bytes)):
channels = [channels]
channels = [self._encoder(x) for x in channels]
if patterns:
if isinstance(patterns, (str, bytes)):
patterns = [patterns]
patterns = [self._encoder(x) for x in patterns]
await cmd(self, channels, patterns)
def _add_message(self, msg):
if self._messages is not None:
self._messages.append(msg)
self._event.set()
def _get_message(self):
try:
msg = self._messages.popleft()
except IndexError:
return None
if self._decoder:
msg = self._decoder(msg)
if isinstance(msg, Exception):
raise msg
return msg