def test_time_diff(self):
"""Test time_diff."""
self.assertEqual(
Timing.time_diff("20180106_120000", "20180106_120010"),
datetime.timedelta(0, 10))
self.assertEqual(
Timing.time_diff("20180106_110000", "20180106_120010"),
datetime.timedelta(0, 3610))
def get_notification_message(self, is_full, start_timestamp, end_timestamp,
success, blob_size, blob_path, error_msg):
"""Assemble JSON message for notification."""
data = {
"cloud":
"azure",
"hostname":
self.backup_configuration.get_vm_name(),
"instance-id":
self.backup_configuration.get_system_uuid(),
"state": {
True: "success",
False: "fail"
}[success],
"type":
"fs",
"method":
"file",
"level": {
True: "full",
False: "incr"
}[is_full],
"account-id":
self.backup_configuration.get_subscription_id(),
"customer-id":
self.backup_configuration.cfg_file_value("DEFAULT.CID"),
"system-id":
self.backup_configuration.cfg_file_value("DEFAULT.SID"),
"database-name":
"",
"database-id":
"",
"s3-path":
self.backup_configuration.get_azure_storage_account_name() +
'.blob.core.windows.net' + blob_path,
"timestamp-send":
Timing.local_string_to_utc_epoch(Timing.now_localtime()),
"timestamp-last-successful":
Timing.local_string_to_utc_epoch(start_timestamp),
"timestamp-bkp-begin":
Timing.local_string_to_utc_epoch(start_timestamp),
"timestamp-bkp-end":
Timing.local_string_to_utc_epoch(end_timestamp),
"backup-size":
blob_size,
"dbtype":
"",
"error-message":
error_msg or '',
"script-version":
azfilebak.__version__
}
return json.dumps(data)
def latest_backup_timestamp(self, fileset, is_full):
"""Return the timestamp for the latest backup for a given fileset."""
existing_blobs_dict = self.existing_backups_for_fileset(
fileset=fileset, is_full=is_full)
if not existing_blobs_dict.keys():
return "19000101_000000"
return Timing.sort(existing_blobs_dict.keys())[-1:][0]
def test_parse(self):
"""Test parse."""
res = Timing.parse("20180605_215959")
self.assertEqual(
time.struct_time(
(2018, 6, 5, 21, 59, 59, 1, 156, -1)),
res)
def test_sort(self):
"""Test sort."""
self.assertEqual(
Timing.sort(['20180110_120000', '20180105_120000', '20180101_120000']),
['20180101_120000', '20180105_120000', '20180110_120000'])
self.assertEqual(
Timing.sort(
['20180105_120000', '20180110_120000', '20180105_120000', '20180101_120000']),
['20180101_120000', '20180105_120000', '20180105_120000', '20180110_120000'])
pick_start_date = lambda x: x["start_date"]
self.assertEqual(
Timing.sort(
times=self.__recovery_sample_data(),
selector=pick_start_date),
[
{'is_full': True, 'start_date': '20180101_010000'},
{'is_full': True, 'start_date': '20180101_010000'},
{'is_full': True, 'start_date': '20180101_010000'},
{'is_full': False, 'start_date': '20180101_011000'},
{'is_full': False, 'start_date': '20180101_012000'},
{'is_full': False, 'start_date': '20180101_013000'},
{'is_full': True, 'start_date': '20180101_014000'},
{'is_full': True, 'start_date': '20180101_014000'},
{'is_full': True, 'start_date': '20180101_014000'},
{'is_full': False, 'start_date': '20180101_015000'},
{'is_full': False, 'start_date': '20180101_020000'},
{'is_full': False, 'start_date': '20180101_021000'},
{'is_full': False, 'start_date': '20180101_021000'},
{'is_full': True, 'start_date': '20180101_022000'},
{'is_full': True, 'start_date': '20180101_022000'},
{'is_full': True, 'start_date': '20180101_022000'},
{'is_full': False, 'start_date': '20180101_023000'},
{'is_full': False, 'start_date': '20180101_024000'},
{'is_full': False, 'start_date': '20180101_025000'},
{'is_full': True, 'start_date': '20180101_030000'},
{'is_full': True, 'start_date': '20180101_030000'},
{'is_full': True, 'start_date': '20180101_030000'},
{'is_full': False, 'start_date': '20180101_031000'},
{'is_full': False, 'start_date': '20180101_032000'},
{'is_full': False, 'start_date': '20180101_032000'},
{'is_full': False, 'start_date': '20180101_033000'}
])
self.assertEquals(
map(pick_start_date, Timing.sort(times=self.__recovery_sample_data(), selector=pick_start_date)),
['20180101_010000', '20180101_010000', '20180101_010000', '20180101_011000', '20180101_012000', '20180101_013000', '20180101_014000', '20180101_014000', '20180101_014000', '20180101_015000', '20180101_020000', '20180101_021000', '20180101_021000', '20180101_022000', '20180101_022000', '20180101_022000', '20180101_023000', '20180101_024000', '20180101_025000', '20180101_030000', '20180101_030000', '20180101_030000', '20180101_031000', '20180101_032000', '20180101_032000', '20180101_033000'])
def prune_old_backups(self, older_than, filesets):
"""
Delete (prune) old backups from Azure storage.
"""
minimum_deletable_age = datetime.timedelta(7, 0)
logging.warn("Deleting files older than %s", older_than)
if older_than < minimum_deletable_age:
msg = "Will not delete files younger than {}, ignoring".format(
minimum_deletable_age)
logging.warn(msg)
return
marker = None
while True:
results = self.backup_configuration.storage_client.list_blobs(
container_name=self.backup_configuration.
azure_storage_container_name,
marker=marker)
for blob in results:
parts = Naming.parse_blobname(blob.name)
if parts is None:
continue
(fileset, _is_full, start_timestamp, _vmname) = parts
if (fileset != None) and not fileset in filesets:
continue
diff = Timing.time_diff(start_timestamp,
Timing.now_localtime())
delete = diff > older_than
if delete:
logging.warn("Deleting %s", blob.name)
self.backup_configuration.storage_client.delete_blob(
container_name=self.backup_configuration.
azure_storage_container_name,
blob_name=blob.name)
else:
logging.warn("Keeping %s", blob.name)
if results.next_marker:
marker = results.next_marker
else:
break
def should_run_tran_backup(now_time, force, latest_tran_backup_timestamp,
log_backup_interval_min):
"""Determine if a 'tran' backup can be performed according to backup window rules."""
if force:
return True
age_of_latest_backup_in_storage = Timing.time_diff(
latest_tran_backup_timestamp, now_time)
min_interval_allows_backup = age_of_latest_backup_in_storage > log_backup_interval_min
perform_tran_backup = min_interval_allows_backup
return perform_tran_backup
def is_backup_allowed_time(self, time):
"""
>>> sample_data = BusinessHours._BusinessHours__sample_data()
>>> sample_hours = BusinessHours.parse_tag_str(sample_data)
>>> some_tuesday_evening = "20180605_215959"
>>> sample_hours.is_backup_allowed_time(some_tuesday_evening)
True
>>> some_tuesday_noon = "20180605_115500"
>>> sample_hours.is_backup_allowed_time(some_tuesday_noon)
False
>>> some_sunday_noon = "20180610_115500"
>>> sample_hours.is_backup_allowed_time(some_sunday_noon)
True
"""
# time.struct_time.tm_wday is range [0, 6], Monday is 0
t = Timing.parse(time)
return self.is_backup_allowed_dh(day=1 + t.tm_wday, hour=t.tm_hour)
def should_run_full_backup(now_time, force, latest_full_backup_timestamp,
business_hours, db_backup_interval_min,
db_backup_interval_max):
"""
Determine whether a backup should be executed.
"""
allowed_by_business = business_hours.is_backup_allowed_time(now_time)
age_of_latest_backup_in_storage = Timing.time_diff(
latest_full_backup_timestamp, now_time)
min_interval_allows_backup = age_of_latest_backup_in_storage > db_backup_interval_min
max_interval_requires_backup = age_of_latest_backup_in_storage > db_backup_interval_max
perform_full_backup = (allowed_by_business
and min_interval_allows_backup
or max_interval_requires_backup or force)
# logging.info("Full backup requested. Current time: {now}. Last backup in storage: {last}. Age of backup {age}".format(now=now_time, last=latest_full_backup_timestamp, age=age_of_latest_backup_in_storage))
# logging.info("Backup requirements: min=\"{min}\" max=\"{max}\"".format(min=db_backup_interval_min,max=db_backup_interval_max))
# logging.info("Forced by user: {force}. Backup allowed by business hours: {allowed_by_business}. min_interval_allows_backup={min_interval_allows_backup}. max_interval_requires_backup={max_interval_requires_backup}".format(force=force, allowed_by_business=allowed_by_business, min_interval_allows_backup=min_interval_allows_backup, max_interval_requires_backup=max_interval_requires_backup))
# logging.info("Decision to backup: {perform_full_backup}.".format(perform_full_backup=perform_full_backup))
return perform_full_backup
def backup_single_fileset(self,
fileset,
is_full,
force,
command=None,
rate=None):
"""
Backup a single fileset using the specified command.
If no command is provided, it will be looked up in the config file.
"""
logging.info("Backup request for fileset: %s", fileset)
# Determine if backup can run according to schedule
start_timestamp = Timing.now_localtime()
end_timestamp = None
if not self.should_run_backup(fileset=fileset,
is_full=is_full,
force=force,
start_timestamp=start_timestamp):
logging.warn("Skipping backup of fileset %s", fileset)
return
# Final destination container
dest_container_name = self.backup_configuration.azure_storage_container_name
vmname = self.backup_configuration.get_vm_name()
# Name of the backup blob
blob_name = Naming.construct_blobname(fileset=fileset,
is_full=is_full,
start_timestamp=start_timestamp,
vmname=vmname)
# Command to run to execute the backup
if not command:
command = self.backup_configuration.get_backup_command(fileset)
try:
# Run the backup command
proc = self.executable_connector.run_backup_command(command, rate)
logging.info("Streaming backup to blob: %s in container: %s",
blob_name, dest_container_name)
# Stream backup command stdout to the blob
storage_client = self.backup_configuration.storage_client
storage_client.create_blob_from_stream(
container_name=dest_container_name,
blob_name=blob_name,
stream=proc.stdout,
use_byte_buffer=True,
max_connections=1)
# Wait for the command to terminate
retcode = proc.wait()
# Check return code
# Ignore return code 1 (files changed during backup)
if retcode == 1:
logging.warning("ignoring tar command return code 1")
elif retcode != 0:
raise BackupException(
"tar command failed with return code {}".format(retcode))
except Exception as ex:
logging.error("Failed to stream blob: %s", ex.message)
end_timestamp = Timing.now_localtime()
self.send_notification(is_full=is_full,
start_timestamp=start_timestamp,
end_timestamp=end_timestamp,
success=False,
blob_size=0,
blob_path='/' + dest_container_name + '/' +
blob_name,
error_msg=ex.message)
raise ex
logging.info("Finished streaming blob: %s", blob_name)
end_timestamp = Timing.now_localtime()
# Get blob size
try:
blob_props = storage_client.get_blob_properties(
dest_container_name, blob_name)
except Exception as ex:
logging.error("Failed to get blob size: %s", ex.message)
self.send_notification(is_full=is_full,
start_timestamp=start_timestamp,
end_timestamp=end_timestamp,
success=False,
blob_size=0,
blob_path='/' + dest_container_name + '/' +
blob_name,
error_msg=ex.message)
raise ex
# Send notification
self.send_notification(is_full=is_full,
start_timestamp=start_timestamp,
end_timestamp=end_timestamp,
success=True,
blob_size=blob_props.properties.content_length,
blob_path='/' + dest_container_name + '/' +
blob_name,
error_msg=None)
# Return name of new blob
return blob_name
def is_backup_allowed_now_localtime(self):
return self.is_backup_allowed_time(time=Timing.now_localtime())