很多线上服务器为了提供文件系统IO性能,会在挂载文件系统的时候指定“noatime,nodiratime”参数,意味着当访问一个文件和目录的时候,access time都不会更新。但是如果未指定上面的参数,atime则会更新。那么具体差异在哪里?
未指定 noatime,nodiratime
$ touch test ; stat test ; ... Access: 2015-04-04 00:37:23.507135507 +0800 Modify: 2015-04-04 00:37:23.507135507 +0800 Change: 2015-04-04 00:37:23.507135507 +0800 $ echo hello >> test ; stat test; ... Access: 2015-04-04 00:37:23.507135507 +0800 Modify: 2015-04-04 00:37:38.018430637 +0800 Change: 2015-04-04 00:37:38.018430637 +0800 $ cat test ;stat test ... Access: 2015-04-04 00:38:02.916135510 +0800 Modify: 2015-04-04 00:37:38.018430637 +0800 Change: 2015-04-04 00:37:38.018430637 +0800
可以看出未指定"noatime,nodiratime"的情况下
- read文件的时候会导致atime更新,不会导致mtime和ctime更新
- write文件只会导致mtime和ctime更新,不会导致atime更新。
指定 noatime,nodiratime
$touch test ; stat test ; ... Access: 2015-04-04 00:28:28.680135484 +0800 Modify: 2015-04-04 00:28:28.680135484 +0800 Change: 2015-04-04 00:28:28.680135484 +0800 $ sleep 10 ; echo hello >> test ; stat test; ... Access: 2015-04-04 00:28:28.680135484 +0800 Modify: 2015-04-04 00:28:38.682727983 +0800 Change: 2015-04-04 00:28:38.682727983 +0800 $ cat test ;stat test ... Access: 2015-04-04 00:28:28.680135484 +0800 Modify: 2015-04-04 00:28:38.682727983 +0800 Change: 2015-04-04 00:28:38.682727983 +0800
可以看出指定"noatime,nodiratime"的情况下
- read文件的时候不会导致atime、mtime、ctime改变
- write文件只会导致mtime和ctime更新,不会导致atime更新。
实际应用场景
在平日里经常有删除文件的需求,大概如下
删除过去N天内都未访问过的文件或者目录(删除N天前访问过的文件) $ #注意这条命令很危险! $ find /home/fire/ -atime +N -exec rm -rf {} \;
假设 /home/fire 目录是一周之前创建的,那么对于这条命令有两个执行结果
$ #注意这条命令很危险! $ find /home/fire/ -atime +7 -exec rm -rf {} \;
- 指定"noatime":find的时候发现 /home/fire 是7天之前创建的,立马就会删除整个目录。而且还会报错"find: /home/fire: No such file or directory",原因就是第一个rm -rf /home/fire 之后 find失败了。这种是很危险的!原因是会误删除文件。
- 未指定"noatime":那就得看情况,如果/home/fire过去7天没有被访问过,那么就和情况一一样,直接删除。如果过去7天内,该目录有人访问过,atime肯定是7天之内,那么就会遍历下面的目录,依次按照之前逻辑。但是遍历过程会更改目录的atime。
看了上面的例子会发现find去删除目录的时候变得好复杂,而且一定要小心。所以find删除更适用于删除文件,不要删除目录。
删除N天内未被访问过的文件 $ find /home/fire/ -atime +N -type f -exec rm -f {} \;
注:
上面是转载的原文,基本没有问题,但关于禁止文件和目录的访问时间,只需指定noatime即可,因为在linux内核中,当指定了noatime,关于nodiratime的判断也就自动跳过了。
下面是最新稳定版linux 4.13.3内核 中关于这段的代码说明:
bool __atime_needs_update(const struct path *path, struct inode *inode, bool rcu) { struct vfsmount *mnt = path->mnt; struct timespec now; if (inode->i_flags & S_NOATIME) return false; /* Atime updates will likely cause i_uid and i_gid to be written * back improprely if their true value is unknown to the vfs. */ if (HAS_UNMAPPED_ID(inode)) return false; if (IS_NOATIME(inode)) return false; if ((inode->i_sb->s_flags & MS_NODIRATIME) && S_ISDIR(inode->i_mode)) return false; if (mnt->mnt_flags & MNT_NOATIME) return false; if ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)) return false; now = current_time(inode); if (!relatime_need_update(path, inode, now, rcu)) return false; if (timespec_equal(&inode->i_atime, &now)) return false; return true; }
生产环境中的Linux内核多没有这么新,一般是2.6.32左右的,下面是2.6.32的linux内核关于这一点的说明:
void touch_atime(struct vfsmount *mnt, struct dentry *dentry) { struct inode *inode = dentry->d_inode; struct timespec now; if (inode->i_flags & S_NOATIME) return; if (IS_NOATIME(inode)) return; if ((inode->i_sb->s_flags & MS_NODIRATIME) && S_ISDIR(inode->i_mode)) return; if (mnt->mnt_flags & MNT_NOATIME) return; if ((mnt->mnt_flags & MNT_NODIRATIME) && S_ISDIR(inode->i_mode)) return; now = current_fs_time(inode->i_sb); if (!relatime_need_update(mnt, inode, now)) return; if (timespec_equal(&inode->i_atime, &now)) return; if (mnt_want_write(mnt)) return; inode->i_atime = now; mark_inode_dirty_sync(inode); mnt_drop_write(mnt); } EXPORT_SYMBOL(touch_atime);
通过查看mount(2)可以看到几个参数说明:
MS_NOATIME Do not update access times for (all types of) files on this file system MS_NODIRATIME Do not update access times for directories on this file system. This flag provides a subset of the functionality provided by MS_NOAT- IME; that is, MS_NOATIME implies MS_NODIRATIME. MS_RELATIME (Since Linux 2.6.20) When a file on this file system is accessed, only update the file’s last access time (atime) if the current value of atime is less than or equal to the file’s last modification time (mtime) or last status change time (ctime). This option is useful for programs, such as mutt(1), that need to know when a file has been read since it was last modified. Since Linux 2.6.30, the kernel defaults to the behav- ior provided by this flag (unless MS_NOATIME was specified), and the MS_STRICTATIME flag is required to obtain traditional semantics. In addition, since Linux 2.6.30, the file’s last access time is always updated if it is more than 1 day old. MS_STRICTATIME (Since Linux 2.6.30) Always update the last access time (atime) when files on this file system are accessed. (This was the default behavior before Linux 2.6.30.) Specifying this flag overrides the effect of setting the MS_NOATIME and MS_RELATIME flags.
通过判定MS_NODIRATIME,如果为真,就不再更新目录的访问时间。
上面参数中提到了MS_RELATIME,对于某些用户某些应用,atime 信息可能是必要的,比如邮件程序和备份工具。如果完全禁止 atime 更新会使这些程序无法正常工作。针对这种情况,Linux 在内核 2.6.20 中添加了一个新的 mount 选项 relatime(relative atime 的意思)。relatime 的意思是访问文件时,仅在 atime 早于文件的更改时间时对 atime 进行更新。在内核 2.6.24 中,又对 relatime 进行了扩展,在访问文件时,当 atime 已经超过某个时间(例如一天)没有更新,就对 atime 进行更新。这个扩展的意思就是调整 atime 的更新粒度。
下面是mount(8)中的参数说明:
relatime Update inode access times relative to modify or change time. Access time is only updated if the previous access time was earlier than the current modify or change time. (Similar to noatime, but doesn’t break mutt or other applications that need to know if a file has been read since the last time it was modified.) Since Linux 2.6.30, the kernel defaults to the behavior provided by this option (unless noatime was specified), and the strictatime option is required to obtain traditional semantics. In addition, since Linux 2.6.30, the file’s last access time is always updated if it is more than 1 day old. strictatime Allows to explicitly requesting full atime updates. This makes it possible for kernel to defaults to relatime or noatime but still allow userspace to override it. For more details about the default system mount options see /proc/mounts.
使用 relatime(relative atime)挂载,可在前一个 atime 更新比 mtime 或者 ctime 更新旧时更新 atime。
在这个示例中,GFS2 文件系统位于 /dev/vg01/lvol0,并挂载到目录 /mygfs2。只有在前一个 atime 更新比 mtime 或者 ctime 更新旧时才会进行 atime 更新。
mount /dev/vg01/lvol0 /mygfs2 -o relatime
参考资料:
http://www.firefoxbug.com/index.php/archives/2801/
http://www.cnblogs.com/allegro/archive/2011/04/18/2019598.html
https://mp.weixin.qq.com/s?__biz=MzA3MzYwNjQ3NA==&mid=400742435&idx=1&sn=d5fa0776ebce30be1de897cd7d1eb295&scene=1&srcid=1123K3Ka8EtwjcNXhNfUGJV7#rd
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/tree/fs/inode.c?h=v4.13.3
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/tree/fs/inode.c?h=linux-2.6.32.y