[linux-lvm] Replace Drive in RAID6
Adam Puleo
adam.puleo at icloud.com
Sat Nov 20 07:15:24 UTC 2021
Hello Heinz,
This was CenOS 7.x using a 5.x kernel.
I could not get the LV activated until I mapped lv_data_rmeta_0 & lv_data_rimage_0 volumes to a PV.
I did the mapping by creating a copy of the current VG config file from /etc/lvm/backup and manually adding the missing PV to each of the LVs (lv_data_rmeta_0 & lv_data_rimage_0). After I edited the file I ran vgcfgrestore.
At this point I was able to activate the LV and sync the RAID. After the synchronization was done I mounted the file system (XFS) but it was corrupted. It was missing most of my newer files. I had a backup so I restored from that.
-adam
> On Nov 16, 2021, at 10:46 AM, Heinz Mauelshagen <heinzm at redhat.com> wrote:
>
> Erm, s/lvchange --repair/lvconvert --repair/
>
> On Tue, Nov 16, 2021 at 6:41 PM Heinz Mauelshagen <heinzm at redhat.com <mailto:heinzm at redhat.com>> wrote:
> On Mon, Nov 8, 2021 at 8:08 AM Adam Puleo <adam.puleo at icloud.com <mailto:adam.puleo at icloud.com>> wrote:
> Hello Everyone,
>
> Hi,
> for starters, which kernel/distro is this?
>
> Also, all layout changes on RaidLVs require activated ones.
>
>
>
> Since the sub-LV #0 has errored, LVM will not let me activate the logical volume.
>
> Is there a way to remap the #0 sub-LV to the replaced disk or resizing the RAID6 to one less disk?
>
> Each of the RAID6 SubLV pairs has an internal id and all data and parity (P+Q syndromes) have been stored in a rotating pattern > no to the remapping part.
>
> Also no to the resize, as it'd need a fully operational radi6, thus repairing the RaidLV is needed.
>
> As mentioned, '"lvchange --rebuild ..." is inadequate to repair RaidLVs with broken/lost PVs, "lvchange --repair $RaidLV" is.
>
> In order to diagnose why your raid6 LV now fails to activate via "lvchange -ay --activationmode degraded $RaidLV" which is the proper way to go about it,
> can you please describe any/all updating steps you took after the drive failed rendering your raid6 LV degraded? Please don't change anything until you made
> that transparent so that we keep chances to fix this...
>
> FYI:
> "lvconvert --(repair|replace) ..." difference is the former repairing RaidLVs with failed PVs by allocating space on different, accessible PVs hence causing the RaidLV to become fully operational after rebuilding all missing block content by using parity stored on the remaining rimage SubLVs vs. the latter allowing to replace mappings to intact PVs by remapping the RAID SubLV pair to different ones (e.g. faster or less contended PVs).
>
> Thanks,
> Heinz
>
>
>
> Thank you,
> -Adam
>
>
> On Nov 4, 2021, at 7:21 PM, Adam Puleo <adam.puleo at icloud.com <mailto:adam.puleo at icloud.com>> wrote:
>
> Hello Andreas,
>
> After deactivating each of the individual rmage and rmeta volumes I receive this error:
> # lvchange -a y --activationmode degraded vg_data/lv_data
> device-mapper: reload ioctl on (253:12) failed: Invalid argument
>
> In messages I see the following errors:
> Nov 4 19:19:43 nas kernel: device-mapper: raid: Failed to read superblock of device at position 0
> Nov 4 19:19:43 nas kernel: device-mapper: raid: New device injected into existing raid set without 'delta_disks' or 'rebuild' parameter specified
> Nov 4 19:19:43 nas kernel: device-mapper: table: 253:12: raid: Unable to assemble array: Invalid superblocks
> Nov 4 19:19:43 nas kernel: device-mapper: ioctl: error adding target to table
>
> Am I not adding the new drive to the RAID correctly? I first did a pvcreate and then a vgextend.
>
> I was using the —rebuild option because I know which physical drive is bad. In the lvmraid man page it says —repair might not know which is the correct block to use so to use —rebuild.
>
> Thank you,
> -Adam
>
>
>
> On Nov 3, 2021, at 7:25 AM, Andreas Schrägle <linux-lvm at ajs124.de <mailto:linux-lvm at ajs124.de>> wrote:
>
> On Tue, 2 Nov 2021 22:56:18 -0700
> Adam Puleo <adam.puleo at icloud.com <mailto:adam.puleo at icloud.com>> wrote:
>
> > Hello,
> >
> > One of my drives failed in my RAID6 and I’m trying to replace it without success.
> >
> > I’m trying to rebuild the failed drive (/dev/sda): lvchange --rebuild /dev/sda vg_data
> >
> > But I’m receiving the error: vg_data/lv_data must be active to perform this operation.
> >
> > I have tried to activate the logical volume without success.
> >
> > How do I go about activating the volume so that I can rebuild the failed drive?
> >
> > Thanks,
> > -Adam
> >
> > # lvs -a -o name,segtype,devices
> > LV Type Devices
> > lv_data raid6 lv_data_rimage_0(0),lv_data_rimage_1(0),lv_data_rimage_2(0),lv_data_rimage_3(0),lv_data_rimage_4(0)
> > [lv_data_rimage_0] error
> > [lv_data_rimage_1] linear /dev/sdc1(1)
> > [lv_data_rimage_2] linear /dev/sdb1(1)
> > [lv_data_rimage_3] linear /dev/sdf1(1)
> > [lv_data_rimage_4] linear /dev/sde1(2)
> > [lv_data_rmeta_0] error
> > [lv_data_rmeta_1] linear /dev/sdc1(0)
> > [lv_data_rmeta_2] linear /dev/sdb1(0)
> > [lv_data_rmeta_3] linear /dev/sdf1(0)
> > [lv_data_rmeta_4] linear /dev/sde1(0)
> >
> > # lvs -a
> > LV VG Attr LSize Pool Origin Data% Meta% Move Log Cpy%Sync Convert
> > lv_data vg_data rwi---r--- 990.00g
> > [lv_data_rimage_0] vg_data vwi-a-r-r- 330.00g
> > [lv_data_rimage_1] vg_data Iwi-a-r-r- 330.00g
> > [lv_data_rimage_2] vg_data Iwi-a-r-r- 330.00g
> > [lv_data_rimage_3] vg_data Iwi-a-r-r- 330.00g
> > [lv_data_rimage_4] vg_data Iwi-a-r-r- 330.00g
> > [lv_data_rmeta_0] vg_data ewi-a-r-r- 4.00m
> > [lv_data_rmeta_1] vg_data ewi-a-r-r- 4.00m
> > [lv_data_rmeta_2] vg_data ewi-a-r-r- 4.00m
> > [lv_data_rmeta_3] vg_data ewi-a-r-r- 4.00m
> > [lv_data_rmeta_4] vg_data ewi-a-r-r- 4.00m
> >
> >
> >
> > _______________________________________________
> > linux-lvm mailing list
> > linux-lvm at redhat.com <mailto:linux-lvm at redhat.com>
> > https://listman.redhat.com/mailman/listinfo/linux-lvm <https://listman.redhat.com/mailman/listinfo/linux-lvm>
> > read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/ <http://tldp.org/HOWTO/LVM-HOWTO/>
>
> Hello Adam,
>
> how exactly have you tried to activate the LV so far?
>
> lvchange with --activationmode degraded should work, no?
>
> Also, are you sure that --rebuild is the correct operation?
>
> man 7 lvmraid suggest you might want --repair or --replace instead.
>
> Best Regards
>
>
> _______________________________________________
> linux-lvm mailing list
> linux-lvm at redhat.com <mailto:linux-lvm at redhat.com>
> https://listman.redhat.com/mailman/listinfo/linux-lvm <https://listman.redhat.com/mailman/listinfo/linux-lvm>
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/ <http://tldp.org/HOWTO/LVM-HOWTO/>
>
>
> _______________________________________________
> linux-lvm mailing list
> linux-lvm at redhat.com <mailto:linux-lvm at redhat.com>
> https://listman.redhat.com/mailman/listinfo/linux-lvm <https://listman.redhat.com/mailman/listinfo/linux-lvm>
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/ <http://tldp.org/HOWTO/LVM-HOWTO/>
>
>
> _______________________________________________
> linux-lvm mailing list
> linux-lvm at redhat.com <mailto:linux-lvm at redhat.com>
> https://listman.redhat.com/mailman/listinfo/linux-lvm <https://listman.redhat.com/mailman/listinfo/linux-lvm>
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/ <http://tldp.org/HOWTO/LVM-HOWTO/>_______________________________________________
> linux-lvm mailing list
> linux-lvm at redhat.com
> https://listman.redhat.com/mailman/listinfo/linux-lvm
> read the LVM HOW-TO at http://tldp.org/HOWTO/LVM-HOWTO/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/linux-lvm/attachments/20211120/2f77bcdf/attachment.htm>
More information about the linux-lvm
mailing list