☰

vps guest 的 fsck() 之后，错误的 inode 重新出现

tag-icon • tag-icon mount fsck inode snapshot vps
vps guest 的 fsck() 之后，错误的 inode 重新出现

我有几个由 virt-resize 创建的 vps guest 虚拟机。我通过创建快照和 rsync 文件内容对来宾进行日常备份。但其中一些没有干净的磁盘映像。

一旦我创建快照并挂载快照逻辑卷：

lvcreate -L5G -s -n guest15_snapshot /dev/vgroup/guest15
mount -o loop,offset=262144 /dev/group/guest15_snapshot /mnt/snapshot_dir

不良 inocdes 可通过以下方式找到/var/log/messages：

Jul  2 11:21:41 server lvm[10346]: Monitoring snapshot group-guest17_snapshot
Jul  2 11:21:41 server kernel: EXT4-fs warning (device loop0): ext4_orphan_get: bad orphan inode 198!  e2fsck was run?
Jul  2 11:21:41 server kernel: ext4_test_bit(bit=197, block=145) = 0
Jul  2 11:21:41 server kernel: inode=(null)
Jul  2 11:21:41 server kernel: EXT4-fs (loop0): recovery complete
Jul  2 11:21:41 server kernel: EXT4-fs (loop0): mounted filesystem with ordered data mode. Opts: 
Jul  2 11:21:42 server lvm[10346]: No longer monitoring snapshot group-guest17_snapshot
Jul  2 11:21:46 server lvm[10346]: Monitoring snapshot group-guest19_snapshot
Jul  2 11:21:46 server kernel: EXT4-fs warning (device loop0): ext4_orphan_get: bad orphan inode 25964!  e2fsck was run?
Jul  2 11:21:46 server kernel: ext4_test_bit(bit=1387, block=148) = 0
Jul  2 11:21:46 server kernel: inode=(null)
Jul  2 11:21:46 server kernel: EXT4-fs (loop0): recovery complete
Jul  2 11:21:46 server kernel: EXT4-fs (loop0): mounted filesystem with ordered data mode. Opts: 
Jul  2 11:22:22 server lvm[10346]: No longer monitoring snapshot guest-vm19_snapshot

我试过guestfish：

guestfish -a /dev/vgroup/guest17
><fs> run
><fs> fsck ext4 /dev/sda1

大多数时候它返回0x1，这意味着错误已修复。

但是一旦我再次挂载图像，坏的索引节点就会再次出现。

如何永久修复坏的 inode？

相关内容