Design question about using OMV VM to create RAID via passthrough disks.

  • I have a question about the design passing through 10 disks to an OpenMediaVault VM using RAID6 on the VM.




    When creating this RAID6 on the OMV VM everything works correctly. Something that got me thinking about long-term recovery was when I had to restore my OMV VM, then manually repassing through my 10 disks, except I forgot to mount 1/10 disks which rendered OMV to see the RAID6 filesystem "Missing". I then fixed it by adding the last disk and OMV was able to show RAID6 filesystem "Online"




    My main question is if I were to lose a disk on the PVE host, how would I be able to bring my filesystem back online?




    Here is output using 'ls -n /dev/disk/by-id' which should show all drives and RAID parts:


    ***Note all 10 disks setup with RAID6 via OMV are (sda to sdj)***


    [ICODE]


    root@px17 ~ # ls -n /dev/disk/by-id

    total 0

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 ata-ST16000NM001J-2TW113_ZR6021LH -> ../../sda

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR6021MK -> ../../sdh

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR6021N5 -> ../../sdf

    lrwxrwxrwx 1 0 0 9 Jul 7 22:44 ata-ST16000NM001J-2TW113_ZR6021RG -> ../../sdj

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 ata-ST16000NM001J-2TW113_ZR6021SL -> ../../sdb

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR6022AT -> ../../sdg

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR7024B1 -> ../../sdc

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR70252Y -> ../../sdd

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR702587 -> ../../sde

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 ata-ST16000NM001J-2TW113_ZR7025HZ -> ../../sdi

    lrwxrwxrwx 1 0 0 10 Jul 7 03:50 dm-name-vg0-root -> ../../dm-0

    lrwxrwxrwx 1 0 0 10 Jul 7 03:50 dm-name-vg0-swap -> ../../dm-1

    lrwxrwxrwx 1 0 0 10 Jul 7 03:50 dm-uuid-LVM-9aFKuMey6ADpOxY8TRtVpjFfXxwlizuPCtwwPMsIXh4mUO3Q85Za6MU2QDJydkER -> ../../dm-0

    lrwxrwxrwx 1 0 0 10 Jul 7 03:50 dm-uuid-LVM-9aFKuMey6ADpOxY8TRtVpjFfXxwlizuPNgsiGfc9IwPmOwm15J8QrchVwPzEUbLe -> ../../dm-1

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 lvm-pv-uuid-LHHqtf-iZtb-1LCE-poY5-39yz-qzws-S3ztY6 -> ../../md1

    lrwxrwxrwx 1 0 0 11 Jul 7 03:50 md-name-city17:128TBR6 -> ../../md127

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 md-name-rescue:0 -> ../../md0

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 md-name-rescue:1 -> ../../md1

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 md-uuid-4693467c:62fec27c:85a32b26:bd907402 -> ../../md0

    lrwxrwxrwx 1 0 0 11 Jul 7 03:50 md-uuid-dc8356d5:81ede174:648f3c7f:8157ab85 -> ../../md127

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 md-uuid-fd7fc965:ca629897:f7fbd288:ba4c4cbb -> ../../md1

    lrwxrwxrwx 1 0 0 13 Jul 7 03:50 nvme-eui.36344630525055630025384500000001 -> ../../nvme1n1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-eui.36344630525055630025384500000001-part1 -> ../../nvme1n1p1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-eui.36344630525055630025384500000001-part2 -> ../../nvme1n1p2

    lrwxrwxrwx 1 0 0 13 Jul 7 03:50 nvme-eui.36344630525055650025384500000001 -> ../../nvme0n1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-eui.36344630525055650025384500000001-part1 -> ../../nvme0n1p1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-eui.36344630525055650025384500000001-part2 -> ../../nvme0n1p2

    lrwxrwxrwx 1 0 0 13 Jul 7 03:50 nvme-SAMSUNG_MZQL2960HCJR-00A07_S64FNE0R505563 -> ../../nvme1n1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-SAMSUNG_MZQL2960HCJR-00A07_S64FNE0R505563-part1 -> ../../nvme1n1p1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-SAMSUNG_MZQL2960HCJR-00A07_S64FNE0R505563-part2 -> ../../nvme1n1p2

    lrwxrwxrwx 1 0 0 13 Jul 7 03:50 nvme-SAMSUNG_MZQL2960HCJR-00A07_S64FNE0R505565 -> ../../nvme0n1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-SAMSUNG_MZQL2960HCJR-00A07_S64FNE0R505565-part1 -> ../../nvme0n1p1

    lrwxrwxrwx 1 0 0 15 Jul 7 03:50 nvme-SAMSUNG_MZQL2960HCJR-00A07_S64FNE0R505565-part2 -> ../../nvme0n1p2

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc26f7b9 -> ../../sdd

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc26f878 -> ../../sdi

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc27013d -> ../../sde

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc2708ac -> ../../sdc

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc2783e8 -> ../../sdh

    lrwxrwxrwx 1 0 0 9 Jul 7 22:44 wwn-0x5000c500dc278490 -> ../../sdj

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 wwn-0x5000c500dc278a7c -> ../../sdb

    lrwxrwxrwx 1 0 0 9 Jul 7 03:50 wwn-0x5000c500dc278b23 -> ../../sda

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc27989a -> ../../sdf

    lrwxrwxrwx 1 0 0 9 Jul 7 04:34 wwn-0x5000c500dc27a271 -> ../../sdg

    [/ICODE]




    Lastly here is a 'blkid' output:


    [ICODE]


    /dev/nvme0n1p1: UUID="4693467c-62fe-c27c-85a3-2b26bd907402" UUID_SUB="b7294d12-8d83-a358-c469-c10e4514423c" LABEL="rescue:0" TYPE="linux_raid_member" PARTUUID="679908e3-01"

    /dev/nvme0n1p2: UUID="fd7fc965-ca62-9897-f7fb-d288ba4c4cbb" UUID_SUB="72e5ba9c-633c-3172-2eee-4f4bb4dc0692" LABEL="rescue:1" TYPE="linux_raid_member" PARTUUID="679908e3-02"

    /dev/nvme1n1p1: UUID="4693467c-62fe-c27c-85a3-2b26bd907402" UUID_SUB="460c24df-42a1-f131-ba12-6439f8583189" LABEL="rescue:0" TYPE="linux_raid_member" PARTUUID="4cd07aa2-01"

    /dev/nvme1n1p2: UUID="fd7fc965-ca62-9897-f7fb-d288ba4c4cbb" UUID_SUB="f30e72e4-cc01-89ce-2980-d30b0a481322" LABEL="rescue:1" TYPE="linux_raid_member" PARTUUID="4cd07aa2-02"

    /dev/md0: UUID="60f46c67-a299-4c26-b247-533a08f5a6ef" BLOCK_SIZE="1024" TYPE="ext3"

    /dev/md1: UUID="LHHqtf-iZtb-1LCE-poY5-39yz-qzws-S3ztY6" TYPE="LVM2_member"

    /dev/sda: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="7f1a41bf-673d-d1bd-8211-c086bf79eb44" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdb: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="c7310212-7c79-ecf3-3a07-cb1c04abcb5e" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdc: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="b3604e36-6e70-0cc5-bc77-c318ba7b1e62" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/md127: LABEL="volcity17" UUID="a6ea8405-dc47-4db9-be90-4b8373bad82c" BLOCK_SIZE="512" TYPE="xfs"

    /dev/sdd: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="6443204d-234b-c158-7a74-6acfeba4a476" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sde: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="c374e7f0-a923-c962-1225-a1edccbf29ee" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdf: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="04a59cca-eb61-864f-9e27-a39fe7815f23" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdg: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="8f6f0668-1767-8aa6-2d2f-d03ff3ab1e22" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdh: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="30713f85-fdb9-f3b1-e457-94fb3147d8ca" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdi: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="8598ffc7-adfa-e505-83fb-009d28c6b9fd" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/sdj: UUID="dc8356d5-81ed-e174-648f-3c7f8157ab85" UUID_SUB="89b14c9a-8842-06b6-81e2-d449098238ed" LABEL="city17:128TBR6" TYPE="linux_raid_member"

    /dev/mapper/vg0-root: UUID="743c26e6-93de-40d9-88b3-52c24342b410" BLOCK_SIZE="4096" TYPE="ext3"

    /dev/mapper/vg0-swap: UUID="5bb0efda-0c3c-4c70-af80-8f65cff99ee9" TYPE="swap

    [/ICODE]




    Should I also mount the /md0, /md1, /dm to my OMV VM in order to fully bring the filesystem on the VM online, for failed RAID drive recovery?




    Sorry, this is loaded or may not even be supported on Proxmox, but I figured I would ask if someone here has a similar setup or situation.




    Thanks![ICODE][/ICODE]

  • I answered my own question by passingthrough the RAID array :

    /sbin/qm set 102 -virtio11 /dev/disk/by-id/md-uuid-dc8356d5:81ede174:648f3c7f:8157ab85;

    This got the RAID filesystem to show online when one of the disks are missing (in the case of a harddrive failure)



    ALSO in order to see the RAID volume under the "RAID MGMT" tab in OMV you must passthrough disks using the 'scsi' passthrough command:

    example: /sbin/qm set 102 -scsi1 /dev/disk/by-id/wwn-0x5000c500dc278b23;

    NOT: /sbin/qm set 102 -virtio1 /dev/disk/by-id/wwn-0x5000c500dc278b23;

  • fixusstreams

    Hat das Label gelöst hinzugefügt.

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!