def run(self):
try:
self.oplogs['backup'] = Oplog(self.mongodump_oplog['file'],
self.do_gzip(), 'a+',
self.flush_docs, self.flush_secs)
self.oplogs['tailed'] = Oplog(self.tailed_oplog['file'],
self.do_gzip())
logging.info("Resolving oplog for %s to max ts: %s" %
(self.uri, self.max_end_ts))
self.state.set('running', True)
self.state.set('first_ts', self.mongodump_oplog['first_ts'])
if not self.state.get('first_ts'):
self.state.set('first_ts', self.tailed_oplog['first_ts'])
for change in decode_file_iter(
self.oplogs['tailed'],
CodecOptions(unicode_decode_error_handler="ignore")):
self.last_ts = change['ts']
if not self.mongodump_oplog[
'last_ts'] or self.last_ts > self.mongodump_oplog[
'last_ts']:
if self.last_ts < self.max_end_ts:
self.oplogs['backup'].add(change)
self.changes += 1
elif self.last_ts > self.max_end_ts:
break
self.state.set('count',
self.mongodump_oplog['count'] + self.changes)
self.state.set('last_ts', self.last_ts)
self.state.set('running', False)
self.exit_code = 0
except Exception, e:
raise Error("Resolving of oplogs failed! Error: %s" % e)
def server_version(self):
status = self.admin_command('serverStatus')
try:
if 'version' in status:
version = status['version'].split('-')[0]
return tuple(version.split('.'))
except Exception, e:
raise Error("Unable to determine version from serverStatus! Error: %s" % e)
def run(self):
try:
thread_count = self.threads()
self._pool = Pool(processes=thread_count)
logging.info("Archiving backup directories with pool of %i thread(s)" % thread_count)
except Exception, e:
logging.fatal("Could not start pool! Error: %s" % e)
raise Error(e)
def __init__(self, config, sharding, db):
self.config = config
self.sharding = sharding
self.db = db
self.max_lag_secs = self.config.replication.max_lag_secs
self.replsets = {}
self.replset_conns = {}
# Check Sharding class:
if not isinstance(self.sharding, Sharding):
raise Error("'sharding' field is not an instance of class: 'Sharding'!")
# Get a DB connection
if isinstance(self.db, DB):
self.connection = self.db.connection()
if not self.db.is_mongos() and not self.db.is_configsvr():
raise Error('MongoDB connection is not to a mongos or configsvr!')
else:
raise Error("'db' field is not an instance of class: 'DB'!")
class Tar(Task):
def __init__(self, manager, config, timer, base_dir, backup_dir, **kwargs):
super(Tar, self).__init__(self.__class__.__name__, manager, config, timer, base_dir, backup_dir, **kwargs)
self.compression_method = self.config.archive.tar.compression
self.binary = "tar"
self._pool = None
self._pooled = []
def done(self, done_dir):
if done_dir in self._pooled:
logging.debug("Archiving completed for: %s" % done_dir)
self._pooled.remove(done_dir)
else:
raise OperationError("Unexpected response from tar thread: %s" % done_dir)
def wait(self):
if len(self._pooled) > 0:
self._pool.close()
while len(self._pooled):
logging.debug("Waiting for %i tar thread(s) to stop" % len(self._pooled))
sleep(2)
self._pool.terminate()
logging.debug("Stopped all tar threads")
self.stopped = True
self.running = False
def run(self):
try:
thread_count = self.threads()
self._pool = Pool(processes=thread_count)
logging.info("Archiving backup directories with pool of %i thread(s)" % thread_count)
except Exception, e:
logging.fatal("Could not start pool! Error: %s" % e)
raise Error(e)
if os.path.isdir(self.backup_dir):
try:
self.running = True
for backup_dir in os.listdir(self.backup_dir):
subdir_name = os.path.join(self.backup_dir, backup_dir)
if not os.path.isdir(os.path.join(subdir_name, "dump")):
continue
output_file = "%s.tar" % subdir_name
if self.do_gzip():
output_file = "%s.tgz" % subdir_name
self._pool.apply_async(TarThread(subdir_name, output_file, self.compression(), self.verbose, self.binary).run, callback=self.done)
self._pooled.append(subdir_name)
except Exception, e:
self._pool.terminate()
logging.fatal("Could not create tar archiving thread! Error: %s" % e)
raise Error(e)
finally:
def run(self, *args):
if self._task and len(self.notifications) > 0:
try:
logging.info("Sending %i notification(s) to: %s" %
(len(self.notifications), self._task.server))
while len(self.notifications) > 0:
try:
(success, message) = self.notifications.pop()
state = self._task.failed
if success is True:
state = self._task.success
self._task.run(state, message)
except NotifyError:
continue
except Exception, e:
raise Error(e)
class Stage(object):
def __init__(self, stage_name, manager, config, timers, base_dir, backup_dir, **kwargs):
self.stage_name = stage_name
self.manager = manager
self.config = config
self.timers = timers
self.base_dir = base_dir
self.backup_dir = backup_dir
self.args = kwargs
self.running = False
self.stopped = False
self.completed = False
self.stage = "mongodb_consistent_backup.%s" % self.stage_name
self.module = None
self.task = "none"
self._task = None
def init(self):
mod_class = None
if self.task == "none":
logging.info("%s stage disabled, skipping" % self.stage_name)
return
try:
module = sys.modules["%s.%s" % (self.stage, self.task.capitalize())]
mod_class = getattr(module, self.task.capitalize())
except LookupError, e:
raise OperationError('Could not load task %s: %s' % (self.task, e))
if mod_class:
self._task = mod_class(
self.manager,
self.config,
self.timers,
self.base_dir,
self.backup_dir,
**self.args
)
if isinstance(self._task, Task):
logging.debug("Loaded stage %s with task %s" % (self.stage, self.task.capitalize()))
else:
raise Error("Loaded class must be child of mongodb_consistent_backup.Pipeline.Task!")
def __init__(self, config, db):
self.config = config
self.db = db
self.read_pref_tags = self.config.replication.read_pref_tags
self.max_lag_secs = self.config.replication.max_lag_secs
self.min_priority = self.config.replication.min_priority
self.max_priority = self.config.replication.max_priority
self.hidden_only = self.config.replication.hidden_only
self.preferred_members = []
if self.config.replication.preferred_members:
self.preferred_members = self.config.replication.preferred_members.split(",")
logging.debug("Preferred members: %s" % self.preferred_members)
self.state_primary = 1
self.state_secondary = 2
self.state_arbiter = 7
self.hidden_weight = 0.20
self.pri0_weight = 0.10
self.replset = True
self.rs_config = None
self.rs_status = None
self.primary = None
self.secondary = None
self.mongo_config = None
self.replset_summary = {}
# Get a DB connection
try:
if isinstance(self.db, DB):
self.connection = self.db.connection()
else:
raise Error("'db' field is not an instance of class: 'DB'!")
except Exception, e:
logging.fatal("Could not get DB connection! Error: %s" % e)
raise OperationError(e)
def close(self, code=None, frame=None):
raise Error("Must define a .close() method when using %s class!" %
self.__class__.__name__)
def run(self):
raise Error("Must define a .run() method when using %s class!" %
self.__class__.__name__)
class Tar(Task):
def __init__(self, manager, config, timer, base_dir, backup_dir, **kwargs):
super(Tar, self).__init__(self.__class__.__name__, manager, config,
timer, base_dir, backup_dir, **kwargs)
self.compression_method = self.config.archive.tar.compression
self.binary = self.config.archive.tar.binary
self._pool = None
self._pooled = []
self.threads(self.config.archive.tar.threads)
self._all_threads_successful = True
def done(self, result):
success = result["success"]
message = result["message"]
error = result["error"]
directory = result["directory"]
exit_code = result["exit_code"]
if success:
if directory in self._pooled:
logging.debug("Archiving completed for: %s" % directory)
else:
logging.warning(
"Tar thread claimed success, but delivered unexpected response %s for directory %s. "
"Assuming failure anyway." % (message, directory))
self._all_threads_successful = False
else:
self._all_threads_successful = False
logging.error(
"Tar thread failed for directory %s: %s; Exit code %s; Error %s)"
% (directory, message, exit_code, error))
self._pooled.remove(directory)
def wait(self):
if len(self._pooled) > 0:
self._pool.close()
while len(self._pooled):
logging.debug("Waiting for %i tar thread(s) to stop" %
len(self._pooled))
sleep(2)
self._pool.terminate()
logging.debug("Stopped all tar threads")
self.stopped = True
self.running = False
def run(self):
try:
self._pool = Pool(processes=self.threads())
logging.info(
"Archiving backup directories with pool of %i thread(s)" %
self.threads())
except Exception, e:
logging.fatal("Could not start pool! Error: %s" % e)
raise Error(e)
if os.path.isdir(self.backup_dir):
try:
self.running = True
for backup_dir in os.listdir(self.backup_dir):
subdir_name = os.path.join(self.backup_dir, backup_dir)
if not os.path.isdir(os.path.join(subdir_name, "dump")):
continue
output_file = "%s.tar" % subdir_name
if self.do_gzip():
output_file = "%s.tgz" % subdir_name
self._pool.apply_async(TarThread(subdir_name, output_file,
self.compression(),
self.verbose,
self.binary).run,
callback=self.done)
self._pooled.append(subdir_name)
except Exception, e:
self._pool.terminate()
logging.fatal(
"Could not create tar archiving thread! Error: %s" % e)
raise Error(e)
finally:
