def desc_files_backup(job_data):
''' Function, creates a desc backup of directories.
At the entrance receives a dictionary with the data of the job.
'''
try:
job_name = job_data['job']
backup_type = job_data['type']
tmp_dir = job_data['tmp_dir']
sources = job_data['sources']
storages = job_data['storages']
except KeyError as e:
log_and_mail.writelog('ERROR', f"Missing required key:'{e}'!",
config.filelog_fd, job_name)
return 1
full_path_tmp_dir = general_function.get_tmp_dir(tmp_dir, backup_type)
for i in range(len(sources)):
exclude_list = sources[i].get('excludes', '')
try:
target_list = sources[i]['target']
gzip = sources[i]['gzip']
except KeyError as e:
log_and_mail.writelog('ERROR', f"Missing required key:'{e}'!",
config.filelog_fd, job_name)
continue
# Keeping an exception list in the global variable due to the specificity of
# the `filter` key of the `add` method of the `tarfile` class
general_files_func.EXCLUDE_FILES = general_files_func.get_exclude_ofs(target_list,
exclude_list)
# The backup name is selected depending on the particular glob patterns from
# the list `target_list`
for regex in target_list:
target_ofs_list = general_files_func.get_ofs(regex)
if not target_ofs_list:
log_and_mail.writelog('ERROR', "No file system objects found that" +\
f"match the regular expression '{regex}'!",
config.filelog_fd, job_name)
continue
for i in target_ofs_list:
# Create a backup only if the directory is not in the exception list
# so as not to generate empty backups
if not general_files_func.is_excluded_ofs(i):
# A function that by regularity returns the name of
# the backup WITHOUT EXTENSION AND DATE
backup_file_name = general_files_func.get_name_files_backup(regex, i)
# Get the part of the backup storage path for this archive relative to
# the backup dir
part_of_dir_path = backup_file_name.replace('___', '/')
backup_full_tmp_path = general_function.get_full_path(
full_path_tmp_dir,
backup_file_name,
'tar',
gzip)
periodic_backup.remove_old_local_file(storages, part_of_dir_path, job_name)
if general_files_func.create_tar('files', backup_full_tmp_path, i,
gzip, backup_type, job_name):
# If the dump collection in the temporary directory has successfully
# transferred the data to the specified storage
periodic_backup.general_desc_iteration(backup_full_tmp_path,
storages, part_of_dir_path,
job_name)
else:
continue
# After all the manipulations, delete the created temporary directory and
# data inside the directory with cache davfs, but not the directory itself!
general_function.del_file_objects(backup_type,
full_path_tmp_dir, '/var/cache/davfs2/*')
def desc_files_backup(job_data):
""" Function, creates a desc backup of directories.
At the entrance receives a dictionary with the data of the job.
"""
is_prams_read, job_name, backup_type, tmp_dir, sources, storages, safety_backup, deferred_copying_level = \
general_function.get_job_parameters(job_data)
if not is_prams_read:
return
full_path_tmp_dir = general_function.get_tmp_dir(tmp_dir, backup_type)
dumped_ofs = {}
for i in range(len(sources)):
exclude_list = sources[i].get('excludes', '')
try:
target_list = sources[i]['target']
gzip = sources[i]['gzip']
except KeyError as e:
log_and_mail.writelog('ERROR', f"Missing required key:'{e}'!",
config.filelog_fd, job_name)
continue
# Keeping an exception list in the global variable due to the specificity of
# the `filter` key of the `add` method of the `tarfile` class
general_files_func.EXCLUDE_FILES = general_files_func.get_exclude_ofs(
target_list, exclude_list)
# The backup name is selected depending on the particular glob patterns from
# the list `target_list`
for regex in target_list:
target_ofs_list = general_files_func.get_ofs(regex)
if not target_ofs_list:
log_and_mail.writelog(
'ERROR', "No file system objects found that" +
f"match the regular expression '{regex}'!",
config.filelog_fd, job_name)
continue
for ofs in target_ofs_list:
# Create a backup only if the directory is not in the exception list
# so as not to generate empty backups
if not general_files_func.is_excluded_ofs(ofs):
# A function that by regularity returns the name of
# the backup WITHOUT EXTENSION AND DATE
backup_file_name = general_files_func.get_name_files_backup(
regex, ofs)
# Get the part of the backup storage path for this archive relative to
# the backup dir
part_of_dir_path = backup_file_name.replace('___', '/')
backup_full_tmp_path = general_function.get_full_path(
full_path_tmp_dir, backup_file_name, 'tar', gzip)
periodic_backup.remove_old_local_file(
storages, part_of_dir_path, job_name)
if general_files_func.create_tar('files',
backup_full_tmp_path, ofs,
gzip, backup_type,
job_name):
dumped_ofs[ofs] = {
'success': True,
'tmp_path': backup_full_tmp_path,
'part_of_dir_path': part_of_dir_path
}
else:
dumped_ofs[ofs] = {'success': False}
if deferred_copying_level <= 0 and dumped_ofs[ofs][
'success']:
periodic_backup.general_desc_iteration(
backup_full_tmp_path, storages, part_of_dir_path,
job_name, safety_backup)
else:
continue
for ofs, result in dumped_ofs.items():
if deferred_copying_level == 1 and result['success']:
periodic_backup.general_desc_iteration(
result['tmp_path'], storages,
result['part_of_dir_path'], job_name, safety_backup)
for ofs, result in dumped_ofs.items():
if deferred_copying_level == 2 and result['success']:
periodic_backup.general_desc_iteration(
result['tmp_path'], storages, result['part_of_dir_path'],
job_name, safety_backup)
for ofs, result in dumped_ofs.items():
if deferred_copying_level >= 3 and result['success']:
periodic_backup.general_desc_iteration(result['tmp_path'],
storages,
result['part_of_dir_path'],
job_name, safety_backup)
# After all the manipulations, delete the created temporary directory and
# data inside the directory with cache davfs, but not the directory itself!
general_function.del_file_objects(backup_type, full_path_tmp_dir,
'/var/cache/davfs2/*')
def create_inc_file(local_dst_dirname, remote_dir, part_of_dir_path,
backup_file_name, target, exclude_list, gzip, job_name,
storage, host, share):
''' The function determines whether to collect a full backup or incremental,
prepares all the necessary information.
'''
date_year = general_function.get_time_now('year')
date_month = general_function.get_time_now('moy')
date_day = general_function.get_time_now('dom')
if int(date_day) < 11:
daily_prefix = 'day_01'
elif int(date_day) < 21:
daily_prefix = 'day_11'
else:
daily_prefix = 'day_21'
year_dir = os.path.join(local_dst_dirname, part_of_dir_path, date_year)
initial_dir = os.path.join(year_dir, 'year') # Path to full backup
month_dir = os.path.join(year_dir, 'month_%s' % (date_month), 'monthly')
daily_dir = os.path.join(year_dir, 'month_%s' % (date_month), 'daily',
daily_prefix)
year_inc_file = os.path.join(initial_dir, 'year.inc')
month_inc_file = os.path.join(month_dir, 'month.inc')
daily_inc_file = os.path.join(daily_dir, 'daily.inc')
link_dict = {} # dict for symlink with pairs like dst: src
copy_dict = {} # dict for copy with pairs like dst: src
# Before we proceed to collect a copy, we need to delete the copies for the same month last year
# if they are to not save extra archives
old_year = int(date_year) - 1
old_year_dir = os.path.join(local_dst_dirname, part_of_dir_path,
str(old_year))
if os.path.isdir(old_year_dir):
old_month_dir = os.path.join(old_year_dir, 'month_%s' % (date_month))
del_old_inc_file(old_year_dir, old_month_dir)
if not os.path.isfile(year_inc_file):
# There is no original index file, so we need to check the existence of an year directory
if os.path.isdir(year_dir):
# There is a directory, but there is no file itself, then something went wrong, so
# we delete this directory with all the data inside, because even if they are there
# continue to collect incremental copies it will not be able to
general_function.del_file_objects(job_name, year_dir)
dirs_for_log = general_function.get_dirs_for_log(
year_dir, remote_dir, storage)
file_for_log = os.path.join(dirs_for_log,
os.path.basename(year_inc_file))
log_and_mail.writelog('ERROR', "The file %s not found, so the directory %s is cleared." +\
"Incremental backup will be reinitialized " %(file_for_log, dirs_for_log),
config.filelog_fd, job_name)
# Initialize the incremental backup, i.e. collect a full copy
dirs_for_log = general_function.get_dirs_for_log(
initial_dir, remote_dir, storage)
general_function.create_dirs(job_name=job_name,
dirs_pairs={initial_dir: dirs_for_log})
# Get the current list of files and write to the year inc file
meta_info = get_index(target, exclude_list)
with open(year_inc_file, "w") as index_file:
json.dump(meta_info, index_file)
full_backup_path = general_function.get_full_path(
initial_dir, backup_file_name, 'tar', gzip)
general_files_func.create_tar('files', full_backup_path, target, gzip,
'inc_files', job_name, remote_dir,
storage, host, share)
# After creating the full copy, you need to make the symlinks for the inc.file and
# the most collected copy in the month directory of the current month
# as well as in the decade directory if it's local, scp the repository and
# copy inc.file for other types of repositories that do not support symlynk.
month_dirs_for_log = general_function.get_dirs_for_log(
month_dir, remote_dir, storage)
daily_dirs_for_log = general_function.get_dirs_for_log(
daily_dir, remote_dir, storage)
general_function.create_dirs(job_name=job_name,
dirs_pairs={
month_dir: month_dirs_for_log,
daily_dir: daily_dirs_for_log
})
if storage in 'local, scp':
link_dict[month_inc_file] = year_inc_file
link_dict[os.path.join(
month_dir,
os.path.basename(full_backup_path))] = full_backup_path
link_dict[daily_inc_file] = year_inc_file
link_dict[os.path.join(
daily_dir,
os.path.basename(full_backup_path))] = full_backup_path
else:
copy_dict[month_inc_file] = year_inc_file
copy_dict[daily_inc_file] = year_inc_file
else:
symlink_dir = ''
if int(date_day) == 1:
# It is necessary to collect monthly incremental backup relative to the year copy
old_meta_info = specific_function.parser_json(year_inc_file)
new_meta_info = get_index(target, exclude_list)
general_inc_backup_dir = month_dir
# It is also necessary to make a symlink for inc files and backups to the directory with the first decade
symlink_dir = daily_dir
general_dirs_for_log = general_function.get_dirs_for_log(
general_inc_backup_dir, remote_dir, storage)
symlink_dirs_for_log = general_function.get_dirs_for_log(
symlink_dir, remote_dir, storage)
general_function.create_dirs(job_name=job_name,
dirs_pairs={
general_inc_backup_dir:
general_dirs_for_log,
symlink_dir: symlink_dirs_for_log
})
with open(month_inc_file, "w") as index_file:
json.dump(new_meta_info, index_file)
elif int(date_day) == 11 or int(date_day) == 21:
# It is necessary to collect a ten-day incremental backup relative to a monthly copy
try:
old_meta_info = specific_function.parser_json(month_inc_file)
except general_function.MyError as e:
log_and_mail.writelog(
'ERROR',
"Couldn't open old month meta info file '%s': %s!" %
(month_inc_file, e), config.filelog_fd, job_name)
return 2
new_meta_info = get_index(target, exclude_list)
general_inc_backup_dir = daily_dir
general_dirs_for_log = general_function.get_dirs_for_log(
general_inc_backup_dir, remote_dir, storage)
general_function.create_dirs(
job_name=job_name,
dirs_pairs={general_inc_backup_dir: general_dirs_for_log})
with open(daily_inc_file, "w") as index_file:
json.dump(new_meta_info, index_file)
else:
# It is necessary to collect a normal daily incremental backup relative to a ten-day copy
try:
old_meta_info = specific_function.parser_json(daily_inc_file)
except general_function.MyError as e:
log_and_mail.writelog(
'ERROR',
"Couldn't open old decade meta info file '%s': %s!" %
(daily_inc_file, e), config.filelog_fd, job_name)
return 2
new_meta_info = get_index(target, exclude_list)
general_inc_backup_dir = daily_dir
general_dirs_for_log = general_function.get_dirs_for_log(
general_inc_backup_dir, remote_dir, storage)
general_function.create_dirs(
job_name=job_name,
dirs_pairs={general_inc_backup_dir: general_dirs_for_log})
# Calculate the difference between the old and new file states
diff_json = compute_diff(new_meta_info, old_meta_info)
inc_backup_path = general_function.get_full_path(
general_inc_backup_dir, backup_file_name, 'tar', gzip)
# Define the list of files that need to be included in the archive
target_change_list = diff_json['modify']
# Form GNU.dumpdir headers
dict_directory = {} # Dict to store pairs like dir:GNU.dumpdir
excludes = r'|'.join([
fnmatch.translate(x)[:-7] for x in general_files_func.EXCLUDE_FILES
]) or r'$.'
for dir_name, dirs, files in os.walk(target):
first_level_files = []
if re.match(excludes, dir_name):
continue
for file in files:
if re.match(excludes, os.path.join(dir_name, file)):
continue
first_level_files.append(file)
first_level_subdirs = dirs
dict_directory[dir_name] = get_gnu_dumpdir_format(
diff_json, dir_name, target, excludes, first_level_subdirs,
first_level_files)
create_inc_tar(inc_backup_path, remote_dir, dict_directory,
target_change_list, gzip, job_name, storage, host,
share)
if symlink_dir:
if storage in 'local, scp':
link_dict[daily_inc_file] = month_inc_file
else:
copy_dict[daily_inc_file] = month_inc_file
if link_dict:
for key in link_dict.keys():
src = link_dict[key]
dst = key
try:
general_function.create_symlink(src, dst)
except general_function.MyError as err:
log_and_mail.writelog(
'ERROR',
"Can't create symlink %s -> %s: %s" % (src, dst, err),
config.filelog_fd, job_name)
if copy_dict:
for key in copy_dict.keys():
src = copy_dict[key]
dst = key
try:
general_function.copy_ofs(src, dst)
except general_function.MyError as err:
log_and_mail.writelog(
'ERROR', "Can't copy %s -> %s: %s" % (src, dst, err),
config.filelog_fd, job_name)
def create_inc_backup(local_dst_dirname, remote_dir, part_of_dir_path, backup_file_name,
target, exclude_list, gzip, job_name, storage, host, share, months_to_store):
""" The function determines whether to collect a full backup or incremental,
prepares all the necessary information.
"""
date_year = general_function.get_time_now('year')
date_month = general_function.get_time_now('moy')
date_day = general_function.get_time_now('dom')
dated_paths = get_dated_paths(local_dst_dirname, part_of_dir_path, date_year, date_month, date_day)
# Before we proceed to collect a copy, we need to delete the copies for the same month last year
# if they are to not save extra archives
old_month_dirs = []
if os.path.isdir(dated_paths['old_year_dir']) or months_to_store < 12:
if months_to_store < 12:
int_date_month = int(date_month)
last_month = int_date_month - months_to_store
if last_month <= 0:
m_range = list(range(last_month+12, 13))
m_range.extend(list(range(1, int_date_month)))
else:
m_range = list(range(last_month, int_date_month))
for i in range(1, 13):
if i not in m_range:
date = str(i).zfill(2)
if i < int(date_month):
year_to_cleanup = dated_paths['year_dir']
else:
year_to_cleanup = dated_paths['old_year_dir']
old_month_dirs.append(os.path.join(year_to_cleanup, f'month_{date}'))
else:
old_month_dirs.append(os.path.join(dated_paths['old_year_dir'], f'month_{date_month}'))
del_old_inc_file(dated_paths['old_year_dir'], old_month_dirs)
link_dict = {} # dict for symlink with pairs like dst: src
copy_dict = {} # dict for copy with pairs like dst: src
# Get the current list of files
new_meta_info = get_index(target, exclude_list)
if not os.path.isfile(dated_paths['year_inc_file']):
# There is no original index file, so we need to check the existence of an year directory
if os.path.isdir(dated_paths['year_dir']):
# There is a directory, but there is no file itself, then something went wrong, so
# we delete this directory with all the data inside, because even if they are there
# continue to collect incremental copies it will not be able to
general_function.del_file_objects(job_name, dated_paths['year_dir'])
dirs_for_log = general_function.get_dirs_for_log(dated_paths['year_dir'], remote_dir, storage)
file_for_log = os.path.join(dirs_for_log, os.path.basename(dated_paths['year_inc_file']))
log_and_mail.writelog('ERROR',
f"The file {file_for_log} not found, so the directory {dirs_for_log} is cleared. "
f"Incremental backup will be reinitialized ",
config.filelog_fd, job_name)
# Initialize the incremental backup, i.e. collect a full copy
remote_dir_for_logs = general_function.get_dirs_for_log(dated_paths['initial_dir'], remote_dir, storage)
general_function.create_dirs(job_name=job_name, dirs_pairs={dated_paths['initial_dir']: remote_dir_for_logs})
write_meta_info(dated_paths['year_inc_file'], new_meta_info)
full_backup_path = general_function.get_full_path(dated_paths['initial_dir'],
backup_file_name,
'tar',
gzip)
general_files_func.create_tar('files', full_backup_path, target,
gzip, 'inc_files', job_name,
remote_dir, storage, host, share)
daily_dirs_remote = general_function.get_dirs_for_log(dated_paths['daily_dir'], remote_dir, storage)
month_dirs_remote = general_function.get_dirs_for_log(dated_paths['month_dir'], remote_dir, storage)
general_function.create_dirs(job_name=job_name, dirs_pairs={dated_paths['daily_dir']: daily_dirs_remote,
dated_paths['month_dir']: month_dirs_remote})
if storage in 'local':
link_dict[dated_paths['month_inc_file']] = dated_paths['year_inc_file']
link_dict[os.path.join(dated_paths['month_dir'], os.path.basename(full_backup_path))] = full_backup_path
link_dict[dated_paths['daily_inc_file']] = dated_paths['year_inc_file']
link_dict[os.path.join(dated_paths['daily_dir'], os.path.basename(full_backup_path))] = full_backup_path
elif storage in 'scp, nfs':
copy_dict[dated_paths['month_inc_file']] = dated_paths['year_inc_file']
link_dict[os.path.join(dated_paths['month_dir'], os.path.basename(full_backup_path))] = \
full_backup_path.replace(local_dst_dirname, remote_dir)
copy_dict[dated_paths['daily_inc_file']] = dated_paths['year_inc_file']
link_dict[os.path.join(dated_paths['daily_dir'], os.path.basename(full_backup_path))] = \
full_backup_path.replace(local_dst_dirname, remote_dir)
else:
copy_dict[dated_paths['month_inc_file']] = dated_paths['year_inc_file']
copy_dict[os.path.join(dated_paths['month_dir'], os.path.basename(full_backup_path))] = full_backup_path
copy_dict[dated_paths['daily_inc_file']] = dated_paths['year_inc_file']
copy_dict[os.path.join(dated_paths['daily_dir'], os.path.basename(full_backup_path))] = full_backup_path
else:
symlink_dir = ''
meta_path = ''
if int(date_day) == 1:
meta_path = dated_paths['month_inc_file']
old_meta_path = dated_paths['year_inc_file']
general_inc_backup_dir = dated_paths['month_dir']
symlink_dir = dated_paths['daily_dir']
elif int(date_day) == 11 or int(date_day) == 21:
meta_path = dated_paths['daily_inc_file']
old_meta_path = dated_paths['month_inc_file']
general_inc_backup_dir = dated_paths['daily_dir']
else:
old_meta_path = dated_paths['daily_inc_file']
general_inc_backup_dir = dated_paths['daily_dir']
try:
old_meta_info = specific_function.parser_json(old_meta_path)
except general_function.MyError as e:
log_and_mail.writelog('ERROR',
f"Couldn't open old meta info file '{old_meta_path}': {e}!",
config.filelog_fd, job_name)
return 2
general_dirs_for_log = general_function.get_dirs_for_log(general_inc_backup_dir, remote_dir, storage)
general_function.create_dirs(job_name=job_name, dirs_pairs={general_inc_backup_dir: general_dirs_for_log})
if meta_path:
write_meta_info(meta_path, new_meta_info)
# Calculate the difference between the old and new file states
diff_json = compute_diff(new_meta_info, old_meta_info)
# Define the list of files that need to be included in the archive
target_change_list = diff_json['modify']
dict_directory = get_dict_directory(target, diff_json)
inc_backup_path = general_function.get_full_path(general_inc_backup_dir, backup_file_name, 'tar', gzip)
create_inc_tar(
inc_backup_path, remote_dir, dict_directory, target_change_list, gzip, job_name, storage, host, share
)
if symlink_dir:
symlink_dirs_for_log = general_function.get_dirs_for_log(symlink_dir, remote_dir, storage)
general_function.create_dirs(job_name=job_name, dirs_pairs={symlink_dir: symlink_dirs_for_log})
if storage in 'local':
link_dict[dated_paths['daily_inc_file']] = dated_paths['month_inc_file']
elif storage in 'scp, nfs':
copy_dict[dated_paths['daily_inc_file'].replace(local_dst_dirname, remote_dir)] = \
dated_paths['month_inc_file'].replace(local_dst_dirname, remote_dir)
else:
copy_dict[dated_paths['daily_inc_file']] = dated_paths['month_inc_file']
create_links_and_copies(link_dict, copy_dict, job_name)
