def backup(self): """ Performs a backup for the server """ _logger.debug("initialising backup information") self.executor.init() backup_info = None try: # Create the BackupInfo object representing the backup backup_info = BackupInfo( self.server, backup_id=datetime.datetime.now().strftime('%Y%m%dT%H%M%S')) backup_info.save() self.backup_cache_add(backup_info) output.info( "Starting backup for server %s in %s", self.config.name, backup_info.get_basebackup_directory()) # Run the pre-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'pre') script.env_from_backup_info(backup_info) script.run() # Run the pre-backup-retry-script if present. retry_script = RetryHookScriptRunner( self, 'backup_retry_script', 'pre') retry_script.env_from_backup_info(backup_info) retry_script.run() # Do the backup using the BackupExecutor self.executor.backup(backup_info) # Compute backup size and fsync it on disk self.backup_fsync_and_set_sizes(backup_info) # Mark the backup as DONE backup_info.set_attribute("status", "DONE") # Use BaseException instead of Exception to catch events like # KeyboardInterrupt (e.g.: CRTL-C) except BaseException, e: msg_lines = str(e).strip().splitlines() if backup_info: # Use only the first line of exception message # in backup_info error field backup_info.set_attribute("status", "FAILED") # If the exception has no attached message use the raw type name if len(msg_lines) == 0: msg_lines = [type(e).__name__] backup_info.set_attribute( "error", "failure %s (%s)" % ( self.executor.current_action, msg_lines[0])) output.error("Backup failed %s.\nDETAILS: %s\n%s", self.executor.current_action, msg_lines[0], '\n'.join(msg_lines[1:]))
else: output.info("Backup end at xlog location: %s (%s, %08X)", backup_info.end_xlog, backup_info.end_wal, backup_info.end_offset) output.info("Backup completed") finally: if backup_info: backup_info.save() # Run the post-backup-retry-script if present. try: retry_script = RetryHookScriptRunner( self, 'backup_retry_script', 'post') retry_script.env_from_backup_info(backup_info) retry_script.run() except AbortedRetryHookScript, e: # Ignore the ABORT_STOP as it is a post-hook operation _logger.warning("Ignoring stop request after receiving " "abort (exit code %d) from post-backup " "retry hook script: %s", e.hook.exit_status, e.hook.script) # Run the post-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'post') script.env_from_backup_info(backup_info) script.run() output.result('backup', backup_info)
def backup(self): """ Performs a backup for the server """ _logger.debug("initialising backup information") self.executor.init() backup_info = None try: # Create the BackupInfo object representing the backup backup_info = BackupInfo( self.server, backup_id=datetime.datetime.now().strftime('%Y%m%dT%H%M%S')) backup_info.save() self.backup_cache_add(backup_info) output.info( "Starting backup using %s method for server %s in %s", self.mode, self.config.name, backup_info.get_basebackup_directory()) # Run the pre-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'pre') script.env_from_backup_info(backup_info) script.run() # Run the pre-backup-retry-script if present. retry_script = RetryHookScriptRunner( self, 'backup_retry_script', 'pre') retry_script.env_from_backup_info(backup_info) retry_script.run() # Do the backup using the BackupExecutor self.executor.backup(backup_info) # Compute backup size and fsync it on disk self.backup_fsync_and_set_sizes(backup_info) # Mark the backup as DONE backup_info.set_attribute("status", "DONE") # Use BaseException instead of Exception to catch events like # KeyboardInterrupt (e.g.: CRTL-C) except BaseException as e: msg_lines = str(e).strip().splitlines() if backup_info: # Use only the first line of exception message # in backup_info error field backup_info.set_attribute("status", "FAILED") # If the exception has no attached message use the raw # type name if len(msg_lines) == 0: msg_lines = [type(e).__name__] backup_info.set_attribute( "error", "failure %s (%s)" % ( self.executor.current_action, msg_lines[0])) output.error("Backup failed %s.\nDETAILS: %s\n%s", self.executor.current_action, msg_lines[0], '\n'.join(msg_lines[1:])) else: output.info("Backup end at xlog location: %s (%s, %08X)", backup_info.end_xlog, backup_info.end_wal, backup_info.end_offset) output.info("Backup completed") # Create a restore point after a backup target_name = 'barman_%s' % backup_info.backup_id self.server.postgres.create_restore_point(target_name) finally: if backup_info: backup_info.save() # Make sure we are not holding any PostgreSQL connection # during the post-backup scripts self.server.close() # Run the post-backup-retry-script if present. try: retry_script = RetryHookScriptRunner( self, 'backup_retry_script', 'post') retry_script.env_from_backup_info(backup_info) retry_script.run() except AbortedRetryHookScript as e: # Ignore the ABORT_STOP as it is a post-hook operation _logger.warning("Ignoring stop request after receiving " "abort (exit code %d) from post-backup " "retry hook script: %s", e.hook.exit_status, e.hook.script) # Run the post-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'post') script.env_from_backup_info(backup_info) script.run() output.result('backup', backup_info)
def backup(self, wait=False, wait_timeout=None): """ Performs a backup for the server :param bool wait: wait for all the required WAL files to be archived :param int|None wait_timeout: :return BackupInfo: the generated BackupInfo """ _logger.debug("initialising backup information") self.executor.init() backup_info = None try: # Create the BackupInfo object representing the backup backup_info = LocalBackupInfo( self.server, backup_id=datetime.datetime.now().strftime('%Y%m%dT%H%M%S')) backup_info.set_attribute('systemid', self.server.systemid) backup_info.save() self.backup_cache_add(backup_info) output.info("Starting backup using %s method for server %s in %s", self.mode, self.config.name, backup_info.get_basebackup_directory()) # Run the pre-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'pre') script.env_from_backup_info(backup_info) script.run() # Run the pre-backup-retry-script if present. retry_script = RetryHookScriptRunner(self, 'backup_retry_script', 'pre') retry_script.env_from_backup_info(backup_info) retry_script.run() # Do the backup using the BackupExecutor self.executor.backup(backup_info) # Create a restore point after a backup target_name = 'barman_%s' % backup_info.backup_id self.server.postgres.create_restore_point(target_name) # Free the Postgres connection self.server.postgres.close() # Compute backup size and fsync it on disk self.backup_fsync_and_set_sizes(backup_info) # Mark the backup as WAITING_FOR_WALS backup_info.set_attribute("status", BackupInfo.WAITING_FOR_WALS) # Use BaseException instead of Exception to catch events like # KeyboardInterrupt (e.g.: CTRL-C) except BaseException as e: msg_lines = force_str(e).strip().splitlines() # If the exception has no attached message use the raw # type name if len(msg_lines) == 0: msg_lines = [type(e).__name__] if backup_info: # Use only the first line of exception message # in backup_info error field backup_info.set_attribute("status", BackupInfo.FAILED) backup_info.set_attribute( "error", "failure %s (%s)" % (self.executor.current_action, msg_lines[0])) output.error("Backup failed %s.\nDETAILS: %s", self.executor.current_action, '\n'.join(msg_lines)) else: output.info("Backup end at LSN: %s (%s, %08X)", backup_info.end_xlog, backup_info.end_wal, backup_info.end_offset) executor = self.executor output.info( "Backup completed (start time: %s, elapsed time: %s)", self.executor.copy_start_time, human_readable_timedelta(datetime.datetime.now() - executor.copy_start_time)) # If requested, wait for end_wal to be archived if wait: try: self.server.wait_for_wal(backup_info.end_wal, wait_timeout) self.check_backup(backup_info) except KeyboardInterrupt: # Ignore CTRL-C pressed while waiting for WAL files output.info( "Got CTRL-C. Continuing without waiting for '%s' " "to be archived", backup_info.end_wal) finally: if backup_info: backup_info.save() # Make sure we are not holding any PostgreSQL connection # during the post-backup scripts self.server.close() # Run the post-backup-retry-script if present. try: retry_script = RetryHookScriptRunner( self, 'backup_retry_script', 'post') retry_script.env_from_backup_info(backup_info) retry_script.run() except AbortedRetryHookScript as e: # Ignore the ABORT_STOP as it is a post-hook operation _logger.warning( "Ignoring stop request after receiving " "abort (exit code %d) from post-backup " "retry hook script: %s", e.hook.exit_status, e.hook.script) # Run the post-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'post') script.env_from_backup_info(backup_info) script.run() output.result('backup', backup_info) return backup_info
def delete_backup(self, backup): """ Delete a backup :param backup: the backup to delete :return bool: True if deleted, False if could not delete the backup """ available_backups = self.get_available_backups( status_filter=(BackupInfo.DONE, )) minimum_redundancy = self.server.config.minimum_redundancy # Honour minimum required redundancy if backup.status == BackupInfo.DONE and \ minimum_redundancy >= len(available_backups): output.warning( "Skipping delete of backup %s for server %s " "due to minimum redundancy requirements " "(minimum redundancy = %s, " "current redundancy = %s)", backup.backup_id, self.config.name, minimum_redundancy, len(available_backups)) return False # Keep track of when the delete operation started. delete_start_time = datetime.datetime.now() # Run the pre_delete_script if present. script = HookScriptRunner(self, 'delete_script', 'pre') script.env_from_backup_info(backup) script.run() # Run the pre_delete_retry_script if present. retry_script = RetryHookScriptRunner(self, 'delete_retry_script', 'pre') retry_script.env_from_backup_info(backup) retry_script.run() output.info("Deleting backup %s for server %s", backup.backup_id, self.config.name) previous_backup = self.get_previous_backup(backup.backup_id) next_backup = self.get_next_backup(backup.backup_id) # Delete all the data contained in the backup try: self.delete_backup_data(backup) except OSError as e: output.error("Failure deleting backup %s for server %s.\n%s", backup.backup_id, self.config.name, e) return False # Check if we are deleting the first available backup if not previous_backup: # In the case of exclusive backup (default), removes any WAL # files associated to the backup being deleted. # In the case of concurrent backup, removes only WAL files # prior to the start of the backup being deleted, as they # might be useful to any concurrent backup started immediately # after. remove_until = None # means to remove all WAL files if next_backup: remove_until = next_backup elif BackupOptions.CONCURRENT_BACKUP in self.config.backup_options: remove_until = backup timelines_to_protect = set() # If remove_until is not set there are no backup left if remove_until: # Retrieve the list of extra timelines that contains at least # a backup. On such timelines we don't want to delete any WAL for value in self.get_available_backups( BackupInfo.STATUS_ARCHIVING).values(): # Ignore the backup that is being deleted if value == backup: continue timelines_to_protect.add(value.timeline) # Remove the timeline of `remove_until` from the list. # We have enough information to safely delete unused WAL files # on it. timelines_to_protect -= set([remove_until.timeline]) output.info("Delete associated WAL segments:") for name in self.remove_wal_before_backup(remove_until, timelines_to_protect): output.info("\t%s", name) # As last action, remove the backup directory, # ending the delete operation try: self.delete_basebackup(backup) except OSError as e: output.error( "Failure deleting backup %s for server %s.\n%s\n" "Please manually remove the '%s' directory", backup.backup_id, self.config.name, e, backup.get_basebackup_directory()) return False self.backup_cache_remove(backup) # Save the time of the complete removal of the backup delete_end_time = datetime.datetime.now() output.info( "Deleted backup %s (start time: %s, elapsed time: %s)", backup.backup_id, delete_start_time.ctime(), human_readable_timedelta(delete_end_time - delete_start_time)) # Remove the sync lockfile if exists sync_lock = ServerBackupSyncLock(self.config.barman_lock_directory, self.config.name, backup.backup_id) if os.path.exists(sync_lock.filename): _logger.debug("Deleting backup sync lockfile: %s" % sync_lock.filename) os.unlink(sync_lock.filename) # Run the post_delete_retry_script if present. try: retry_script = RetryHookScriptRunner(self, 'delete_retry_script', 'post') retry_script.env_from_backup_info(backup) retry_script.run() except AbortedRetryHookScript as e: # Ignore the ABORT_STOP as it is a post-hook operation _logger.warning( "Ignoring stop request after receiving " "abort (exit code %d) from post-delete " "retry hook script: %s", e.hook.exit_status, e.hook.script) # Run the post_delete_script if present. script = HookScriptRunner(self, 'delete_script', 'post') script.env_from_backup_info(backup) script.run() return True
def backup(self): """ Performs a backup for the server """ _logger.debug("initialising backup information") self.executor.init() backup_info = None try: # Create the BackupInfo object representing the backup backup_info = BackupInfo( self.server, backup_id=datetime.datetime.now().strftime('%Y%m%dT%H%M%S')) backup_info.save() self.backup_cache_add(backup_info) output.info("Starting backup using %s method for server %s in %s", self.mode, self.config.name, backup_info.get_basebackup_directory()) # Run the pre-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'pre') script.env_from_backup_info(backup_info) script.run() # Run the pre-backup-retry-script if present. retry_script = RetryHookScriptRunner(self, 'backup_retry_script', 'pre') retry_script.env_from_backup_info(backup_info) retry_script.run() # Do the backup using the BackupExecutor self.executor.backup(backup_info) # Compute backup size and fsync it on disk self.backup_fsync_and_set_sizes(backup_info) # Mark the backup as DONE backup_info.set_attribute("status", "DONE") # Use BaseException instead of Exception to catch events like # KeyboardInterrupt (e.g.: CRTL-C) except BaseException as e: msg_lines = str(e).strip().splitlines() if backup_info: # Use only the first line of exception message # in backup_info error field backup_info.set_attribute("status", "FAILED") # If the exception has no attached message use the raw # type name if len(msg_lines) == 0: msg_lines = [type(e).__name__] backup_info.set_attribute( "error", "failure %s (%s)" % (self.executor.current_action, msg_lines[0])) output.error("Backup failed %s.\nDETAILS: %s\n%s", self.executor.current_action, msg_lines[0], '\n'.join(msg_lines[1:])) else: output.info("Backup end at LSN: %s (%s, %08X)", backup_info.end_xlog, backup_info.end_wal, backup_info.end_offset) output.info( "Backup completed (start time: %s, elapsed time: %s)", self.executor.copy_start_time, human_readable_timedelta(self.executor.copy_end_time - self.executor.copy_start_time)) # Create a restore point after a backup target_name = 'barman_%s' % backup_info.backup_id self.server.postgres.create_restore_point(target_name) finally: if backup_info: backup_info.save() # Make sure we are not holding any PostgreSQL connection # during the post-backup scripts self.server.close() # Run the post-backup-retry-script if present. try: retry_script = RetryHookScriptRunner( self, 'backup_retry_script', 'post') retry_script.env_from_backup_info(backup_info) retry_script.run() except AbortedRetryHookScript as e: # Ignore the ABORT_STOP as it is a post-hook operation _logger.warning( "Ignoring stop request after receiving " "abort (exit code %d) from post-backup " "retry hook script: %s", e.hook.exit_status, e.hook.script) # Run the post-backup-script if present. script = HookScriptRunner(self, 'backup_script', 'post') script.env_from_backup_info(backup_info) script.run() output.result('backup', backup_info)