本帖最后由 pig2 于 2014-11-21 15:21 编辑
问题导读
1、审计的主要职责是什么?
2、账户是如何获取全局数据的?
3、如果调用的是AccountAuditor类中的run_forever方法,做什么处理?
概述部分:
账户审计守护进程;
审计的主要的职责就是检查账户数据是否被删除(通过解析账户目录下的.db文件);
这里定义的once=True,说明系统默认调用守护进程类Daemon中的run_once方法,有人说run_once主要应用于测试,这个没有验证;
从而最终实现调用AccountAuditor类中的run_once方法;
如果调用的是AccountAuditor类中的run_forever方法,则会实现循环实现检查指定账户数据是否被删除;
源码解析部分:
- from swift.account.auditor import AccountAuditor
- from swift.common.utils import parse_options
- from swift.common.daemon import run_daemon
- if __name__ == '__main__':
- conf_file, options = parse_options(once=True)
- run_daemon(AccountAuditor, conf_file, **options)
- def run_daemon(klass, conf_file, section_name='', once=False, **kwargs):
- """
- 从配置文件加载设置,然后实例化守护进程“klass”并运行这个守护进程通过指定的参数kwarg;
- """
- ......
- try:
- klass(conf).run(once=once, **kwargs)
- except KeyboardInterrupt:
- logger.info('User quit')
- logger.info('Exited')
- def run(self, once=False, **kwargs):
- """
- Run the daemon
- 运行守护进程程序;
- 即运行方法run_once,或者方法run_forever;
- """
- # 配置参数相关;
- utils.validate_configuration()
- utils.drop_privileges(self.conf.get('user', 'swift'))
- # 日志相关处理;
- utils.capture_stdio(self.logger, **kwargs)
- def kill_children(*args):
- signal.signal(signal.SIGTERM, signal.SIG_IGN)
- os.killpg(0, signal.SIGTERM)
- sys.exit()
- signal.signal(signal.SIGTERM, kill_children)
- if once:
- self.run_once(**kwargs)
- else:
- self.run_forever(**kwargs)
- def run_once(self, *args, **kwargs):
- """
- Override this to run the script once
- 子类中的方法需要被重写;
- """
- raise NotImplementedError('run_once not implemented')
- def run_once(self, *args, **kwargs):
- """
- Run the account audit once.
- 运行一次账户审计;
- 调用方法_one_audit_pass,首先计算获得path,device,partition;
- 然后通过方法account_audit对这些账户内容进行审计,看是否被删除;
- """
- self.logger.info(_('Begin account audit "once" mode'))
- begin = reported = time.time()
- self._one_audit_pass(reported)
- elapsed = time.time() - begin
- self.logger.info(
- _('Account audit "once" mode completed: %.02fs'), elapsed)
- dump_recon_cache({'account_auditor_pass_completed': elapsed},
- self.rcache, self.logger)
- def _one_audit_pass(self, reported):
- """
- 首先通过方法audit_location_generator方法计算获得所有以.db为后缀的文件生成path,device,partition;
- 然后通过方法account_audit对这些账户内容进行审计,看是否被删除;
- """
-
- # 基于给定的设备路径和文件后缀,为在这个数据目录中的所有以.db为后缀的文件生成元组(path, device, partition);
- # 通过方法audit_location_generator方法计算获得path,device,partition;
- all_locs = audit_location_generator(self.devices, DATADIR, '.db',
- mount_check=self.mount_check,
- logger=self.logger)
- for path, device, partition in all_locs:
- # 审计给定的账户,看看是否已经被删除,如果没有删除,则为账户获取全局数据;
- # 获取包括account, created_at, put_timestamp, delete_timestamp, container_count, object_count, bytes_used, hash, id等值的字典;
- self.account_audit(path)
- if time.time() - reported >= 3600: # once an hour
- self.logger.info(_('Since %(time)s: Account audits: '
- '%(passed)s passed audit,'
- '%(failed)s failed audit'),
- {'time': time.ctime(reported),
- 'passed': self.account_passes,
- 'failed': self.account_failures})
- dump_recon_cache({'account_audits_since': reported,
- 'account_audits_passed': self.account_passes,
- 'account_audits_failed':
- self.account_failures},
- self.rcache, self.logger)
- reported = time.time()
- self.account_passes = 0
- self.account_failures = 0
- self.accounts_running_time = ratelimit_sleep(self.accounts_running_time, self.max_accounts_per_second)
复制代码
return reported1.调用方法audit_location_generator,实现基于给定的设备路径(self.devices,DATADIR=accounts)和文件后缀(.db),为在这个数据目录中的所有以.db为后缀的文件生成元组(path, device, partition);
生成路径示例如下:
- path = /devices/device/accounts/partition/asuffix/hsh/****.db
- device:self.devices下具体的一个设备名称;
- partition:/devices/device/accounts/下具体的一个分区名称;
复制代码
2.遍历元组(path, device, partition)中的每一个匹配对,调用方法account_audit实现审计给定的账户,看看是否已经被删除,如果没有删除,则为账户获取全局数据;
转到2,来看方法account_audit:
- def account_audit(self, path):
- """
- 审计给定的账户,看看是否已经被删除,如果没有删除,则为账户获取全局数据;
- 获取包括account, created_at, put_timestamp, delete_timestamp, container_count, object_count, bytes_used, hash, id等值的字典;
- """
- start_time = time.time()
- try:
- broker = AccountBroker(path)
- # is_deleted:检测帐号的数据库是否被删除;
- if not broker.is_deleted():
- # 为账户获取全局数据;
- # 返回包括account, created_at, put_timestamp, delete_timestamp, container_count, object_count, bytes_used, hash, id等值的字典;
- broker.get_info()
- self.logger.increment('passes')
- self.account_passes += 1
- self.logger.debug(_('Audit passed for %s') % broker)
- except (Exception, Timeout):
- self.logger.increment('failures')
- self.account_failures += 1
- self.logger.exception(_('ERROR Could not get account info %s'),
- path)
- self.logger.timing_since('timing', start_time)这个方法的实现比较简单,这里就不再进一步进行解析;
复制代码
|