我通常不会寻求有关 Linux 的帮助，而是喜欢自己解决问题，但这次我真的陷入了困境，我不知道该向谁寻求帮助。
大约一年前，我尝试将 lvmcache 添加到我的一个 LV 中。几天前，我升级了内核（yum update），重启后，我无法再访问该 LV。它上面有我非常想找回的重要数据。
我有两个卷组，SSD 和 Rust：
- SSD 是两个 1Tb NVMe 设备的软件 (MD) raid1 阵列。
- Rust 是一个由四块 10Tb 硬盘组成的软件 (MD) raid5 阵列。
由于各种原因，我不希望 SSD 完全成为缓存而不能用于其他任何用途，因此我在其上设置了 LVM，并为根文件系统、LXC 容器等创建了 LV。这些是我拥有的 LV：
$ sudo lvs
WARNING: Device for PV 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L not found or rejected by a filter.
Couldn't find device with uuid 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L.
LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
Backups Rust -wi-a----- 1.00t
Dxxxxx Rust -wi-a----- 3.00t
Microserver Rust -wi-a----- 5.00t
NAS Rust -wi-a----- 5.00t
Photos Rust Cwi---C-p- 3.00t [Cache] [Photos_corig]
Rescued Rust -wi-a----- 6.00t
Video Rust -wi-a----- 2.00t
CentOS7_Root SSD -wi-ao---- 50.00g
Containers SSD -wi-ao---- 200.00g
MD_Journal SSD -wi-ao---- 16.00g
Rust_Cache SSD -wi-a----- 256.00g
home SSD -wi-ao---- 100.00g
Photos 是无法访问的缓存 LV。设置有点棘手，因为 lvmcache 坚持要求原始 LV 和缓存 LV 位于同一个 VG 中，而我的则不是（一个在 Rust 上，另一个在 SSD 上）。我通过在 SSD 上创建缓存卷，然后将其格式化为 PV，并将该 PV 添加到 Rust 组来解决这个问题。这些是我一年前运行的设置命令：
# Create a cache volume (LV) on the SSD device
lvcreate -L 256G -n Rust_Cache /dev/SSD
# Format that volume as a PV:
pvcreate /dev/SSD/Rust_Cache
# Add it to the Rust VG:
vgextend /dev/Rust /dev/SSD/Rust_Cache
# Create the cache data volume on it:
lvcreate -L 100G -n Cache Rust /dev/SSD/Rust_Cache
# Create the cache meta volume on it:
lvcreate -L 4G -n Cache_Meta Rust /dev/SSD/Rust_Cache
# Combine them into a cache pool:
lvconvert --type cache-pool /dev/Rust/Cache --poolmetadata /dev/Rust/Cache_Meta
# Set that pool as the cache of the Photos LV:
lvconvert --type cache /dev/Rust/Photos --cachepool /dev/Rust/Cache
我认为该问题与上面显示的错误消息有关：
WARNING: Device for PV 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L not found or rejected by a filter.
Couldn't find device with uuid 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L.
这是缓存 PV (/dev/SSD/Rust_Cache) 的 UUID，至少根据/etc/lvm/backup/Rust。我无法确认，因为它拒绝pvdisplay：
$ sudo pvdisplay /dev/SSD/Rust_Cache
Failed to find device for physical volume "/dev/SSD/Rust_Cache".
WARNING: Device for PV 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L not found or rejected by a filter.
Couldn't find device with uuid 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L.
我不认为该设备真的“未找到”，因为 SSD 上的其他卷工作正常，并且它们都共享一个 PV。我认为它一定是被某种“隐藏”LVM 中的缓存卷和原始卷的魔法过滤掉了，但我不知道那是什么。
如果我无法恢复缓存，我准备接受我可能丢失了此卷上的一些数据，但是我不想完全丢失它。希望它全部备份在 S3 上，但考虑到数据量和小文件数量，很难确定。关于我可以做什么来使其再次可访问，有什么建议吗？
vgchange -ay拒绝激活它，但暗示我可以强制它，所以如果无法恢复缓存，这可能是一个选择：
$ sudo vgchange -ay
WARNING: Device for PV 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L not found or rejected by a filter.
Couldn't find device with uuid 0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L.
Refusing activation of partial LV Rust/Photos. Use '--activationmode partial' to override.
6 logical volume(s) in volume group "Rust" now active
5 logical volume(s) in volume group "SSD" now active
blkid根据要求在所有设备上输出：
[sudo] password for chris:
/dev/mapper/SSD-Rust_Cache: UUID="0hAsMD-LsJf-YsiF-iQ0B-tI23-8hso-cRP93L" TYPE="LVM2_member"
/dev/nvme0n1: PTTYPE="gpt"
/dev/nvme0n1p1: UUID="29970b49-140d-6862-786d-f33b5edcab6d" UUID_SUB="7cfe4ba4-3062-3ebb-a2a3-0b3f119e2ccc" LABEL="blackbox.qwarx.com:pv00" TYPE="linux_raid_member" PARTUUID="92abf126-4192-4a7d-8381-85a1ad1b2eaf"
/dev/nvme0n1p2: UUID="a42550e8-adac-3f6e-68ef-41bddb5fa54c" UUID_SUB="cd19178f-2428-2442-0c60-99c4abff3c59" LABEL="blackbox.qwarx.com:Boot" TYPE="linux_raid_member" PARTUUID="3e5c4c02-4483-48fd-82eb-6973bea5674f"
/dev/nvme0n1p3: UUID="1d3395dc-7e7c-bf69-7e51-e49fcb2d085d" UUID_SUB="ede510d6-1468-b9a9-b4af-d8b77af5cd4e" LABEL="blackbox.qwarx.com:EFI" TYPE="linux_raid_member" PARTUUID="81fdb48b-7cd8-4e55-8358-3f3f8aee1800"
/dev/nvme1n1: PTTYPE="gpt"
/dev/nvme1n1p1: UUID="29970b49-140d-6862-786d-f33b5edcab6d" UUID_SUB="dc5e4a8f-3281-cf7e-55a6-b95924872cc6" LABEL="blackbox.qwarx.com:pv00" TYPE="linux_raid_member" PARTUUID="ef1a40a2-c8b2-4428-ad29-fdc1b2640634"
/dev/nvme1n1p2: UUID="a42550e8-adac-3f6e-68ef-41bddb5fa54c" UUID_SUB="54dab930-9e4e-1009-8618-eb100d5de05c" LABEL="blackbox.qwarx.com:Boot" TYPE="linux_raid_member" PARTUUID="26a6252a-871c-4768-916c-1ce038511874"
/dev/nvme1n1p3: UUID="1d3395dc-7e7c-bf69-7e51-e49fcb2d085d" UUID_SUB="e752061b-8410-46fa-c871-04b16b0844bf" LABEL="blackbox.qwarx.com:EFI" TYPE="linux_raid_member" PARTUUID="d16924c1-96b3-4f9f-bef0-8e5ce4402bf0"
/dev/md127: LABEL="Boot" UUID="da2c2ddd-af52-4dfe-92ac-9775cc015234" TYPE="ext4"
/dev/md126: UUID="AKAeOG-fIdp-DKl7-mDiI-LHcH-y09B-3Vxjqw" TYPE="LVM2_member"
/dev/sdb: UUID="653729f4-efb2-af46-dd24-6510380b7c35" UUID_SUB="b2664b6b-fae1-76e1-845e-50faa4dfa13d" LABEL="blackbox.qwarx.com:Rust" TYPE="linux_raid_member"
/dev/sdc: UUID="653729f4-efb2-af46-dd24-6510380b7c35" UUID_SUB="e9defcaf-3cba-2337-17a9-bd247317ecfe" LABEL="blackbox.qwarx.com:Rust" TYPE="linux_raid_member"
/dev/sdd: UUID="653729f4-efb2-af46-dd24-6510380b7c35" UUID_SUB="121f8be1-5ab4-9de0-238d-6ba64aae8c00" LABEL="blackbox.qwarx.com:Rust" TYPE="linux_raid_member"
/dev/sda: UUID="653729f4-efb2-af46-dd24-6510380b7c35" UUID_SUB="dbbffe6d-dae0-31dd-a310-913a1cc8a8e8" LABEL="blackbox.qwarx.com:Rust" TYPE="linux_raid_member"
/dev/mapper/SSD-CentOS7_Root: LABEL="CentOS7_Root" UUID="aa48bf01-c012-4e54-806b-e0a341d548c2" TYPE="xfs"
/dev/md125: SEC_TYPE="msdos" LABEL="EFI" UUID="F61A-994A" TYPE="vfat"
/dev/md124: UUID="RUXdND-W4mD-lXdF-YWMk-J7yY-w50S-zgdxOh" TYPE="LVM2_member"
/dev/mapper/SSD-home: UUID="4031d5d7-dc64-49c4-9ecc-b139709a96ab" TYPE="xfs"
/dev/mapper/SSD-MD_Journal: UUID="653729f4-efb2-af46-dd24-6510380b7c35" UUID_SUB="b504320e-e625-30cd-0270-77b9b63e2482" LABEL="blackbox.qwarx.com:Rust" TYPE="linux_raid_member"
/dev/mapper/SSD-Containers: UUID="c9cd5afb-bdd9-4561-bdac-d3793a7b0c1c" TYPE="ext4"
/dev/mapper/Rust-Rescued: LABEL="Rescued" UUID="9c17d2d8-700c-49c2-98fe-1fb1dde733c6" TYPE="ext4"
/dev/mapper/Rust-Microserver: LABEL="Microserver" UUID="242d2ddf-0759-4346-a545-8df042af5ebe" TYPE="ext4"
/dev/mapper/Rust-NAS: LABEL="NAS" UUID="a37875cf-1f5c-46aa-a825-4204cc98e4c9" TYPE="ext4"
/dev/mapper/Rust-Video: LABEL="Video" UUID="92b9e7e9-874f-433b-aa72-0465546986b7" TYPE="ext4"
/dev/mapper/Rust-Dxxxxx: LABEL="Dxxxxx" UUID="689947c7-7e62-4a3c-a871-59ea923c4dcf" TYPE="ext4"
/dev/mapper/Rust-Backups: LABEL="Backups" UUID="9d0534b2-a545-410f-a9d4-eda4f9836bfb" TYPE="ext4"
答案1
我认为我已经修复它了！
这真的很简单。我知道卷组有一个可用的标志，您可以使用它来更改它：
vgchange -ay /dev/VG
但直到现在我才知道逻辑卷也有这样的标志，并且可以由用户设置。
与卷组一样，似乎有时在不可用的原因消失后，此标志仍会保留。因此，我能够像这样使 Photos LV 可访问，没有任何错误，然后挂载它：
lvchange -ay /dev/Rust/Photos
我确实找到并更改了以下设置/etc/lvm/lvm.conf，因为它似乎可能与此特定设备无法自动激活有关，因此可能已经解决了根本问题/根本原因：
# Configuration option devices/scan_lvs.
# Scan LVM LVs for layered PVs, allowing LVs to be used as PVs.
# When 1, LVM will detect PVs layered on LVs, and caution must be
# taken to avoid a host accessing a layered VG that may not belong
# to it, e.g. from a guest image. This generally requires excluding
# the LVs with device filters. Also, when this setting is enabled,
# every LVM command will scan every active LV on the system (unless
# filtered), which can cause performance problems on systems with
# many active LVs. When this setting is 0, LVM will not detect or
# use PVs that exist on LVs, and will not allow a PV to be created on
# an LV. The LVs are ignored using a built in device filter that
# identifies and excludes LVs.
scan_lvs = 1
答案2
您是否尝试过使用修复盘或 livecd（如 knoppix）进行引导，看看是否可以通过这种方式挂载？如何从 grub 引导菜单引导到以前的内核？