Resize error (too big) 32 bits issue

1. offizieller Beitrag

    Hello,
    I have an openmediavault server, the version is:


    Linux omv 3.16.0-0.bpo.4-amd64 #1 SMP Debian 3.16.39-1+deb8u1~bpo70+1 (2017-02-24) x86_64 GNU/Linux


    The storage setup are 8 disks of 7,3 TB. Initially i created a RAID 6 with 7 disks and then a FS on ext4 with a total space of 36 TB.
    I added then a new 7,3 TB disk to the raid which recognizes correctly, but when i try to expand the FS i get the following error:


    Error #6000:exception 'OMVException' with message 'Failed to grow the filesystem '/dev/md0': resize2fs 1.42.5 (29-Jul-2012)resize2fs: New size too large to be expressed in 32 bits' in /usr/share/openmediavault/engined/rpc/filesystemmgmt.inc:703Stack trace:#0 [internal function]: OMVRpcServiceFileSystemMgmt->resize(Array, Array)#1 /usr/share/php/openmediavault/rpcservice.inc(125): call_user_func_array(Array, Array)#2 /usr/share/php/openmediavault/rpc.inc(79): OMVRpcServiceAbstract->callMethod('resize', Array, Array)#3 /usr/sbin/omv-engined(500): OMVRpc::exec('FileSystemMgmt', 'resize', Array, Array, 1)

    The flag of 64 bit mentioned on another posts is activated.



    I had readed another posts about the 16 Tb problem with the 32 post limit but
    Then, how could i create a 36 TB filesystem with the omw interface when the limit is 16TB?. The problem only comes when i try to resize it?


    My questions:


    Is there any sollution or hotfix to resize this filesystem?



    What happens if i try to create a new FS with more than 36 TB from scratch.


    Thanks.

    • Offizieller Beitrag
    Zitat von jmateos

    Then, how could i create a 36 TB filesystem with the omw interface when the limit is 16TB?. The problem only comes when i try to resize it?

    The OMV web interface doesn't have a 16TB limit. If the 64 bit flag is set, then ext utilities are too old on wheezy. I would upgrade to OMV 3.x or boot a bootable linux distro like systemrescuecd to do the work. The problem with not upgrading is I don't think you could run a filesystem check with the wheezy package.


    Personally, I wouldn't create a 36 TB array. I would create three or four of them and pool them.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    • Offizieller Beitrag
    Zitat von jmateos

    Pool them with what?

    unionfilesystems plugin.


    Zitat von jmateos


    why is better than a 42 Tb filesystem on RAID6?

    Because it doesn't take forever to rebuild if you lose one drive. And that is a lot of data on one filesystem. Even if you have backup, it would take a long ass time to restore. Why do you need a 42 tb filesystem?


    Zitat von jmateos

    There is no posibility to upgrade the related filesystem packages on wheezy?

    You probably could but using packages from non-wheezy sources can cause stability issues. Why not move to omv 3.x?

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    Hello,


    Because the system is working i have users using it and has to be stable, that's why i chose the OMV2 instead of 3. My priority is stability and to be simple. With a RAID6, if i lose a disk it could be replaced without restoring anything or afecting too much the performance, i want system which works with the minimum interaction.


    I saw some problems on the forums related to unionfilesystem plugin, and i'm not sure about to change and get some possible errros.

    • Offizieller Beitrag
    Zitat von jmateos

    Because the system is working i have users using it and has to be stable, that's why i chose the OMV2 instead of 3. My priority is stability and to be simple.

    OMV 3 is stable on my production systems.


    Zitat von jmateos


    With a RAID6, if i lose a disk it could be replaced without restoring anything or afecting too much the performance, i want system which works with the minimum interaction.

    I wasn't suggesting not using raid. I was suggesting using multiple raid arrays pooled together. So, they would be able to do the exact same thing as one array with the same amount of interaction. Replacing a disk and the array rebuilding would definitely affect performance in any configuration.


    Zitat von jmateos

    I saw some problems on the forums related to unionfilesystem plugin, and i'm not sure about to change and get some possible errros.

    I use it on most of my systems. You probably read about errors with aufs or mhddfs or errors in setup. mergerfs is rock solid.


    I was just suggesting other ideas because there is no way I would create a software(mdadm) array that big. If you want one filesystem that big (and better performance), get a real raid controller. Just my opinion...

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    Hello again,


    The configuration that you suggest is to setup 2 or more arrays, and unify on a filesystem with the module. In that case i lose more space on building arrays, (because of the raids) , but i consider the option.


    upgrading from my OMV2 to 3 does not have any risk?, i suppose that the reboot will be needed. do i have any kind of backup to restore the omv2?


    Is there an official document available for the upgrade?


    Once upgraded , i could surely resize my 32 Tb partition ?


    thanks.

    • Offizieller Beitrag
    Zitat von jmateos

    upgrading from my OMV2 to 3 does not have any risk?

    All upgrades have risks. How high depends on what plugins you have installed.



    Zitat von jmateos


    i suppose that the reboot will be needed.

    Yep. You are upgrading every library and the kernel. Won't happen without a reboot.


    Zitat von jmateos

    do i have any kind of backup to restore the omv2?

    Yes. You should have a backup anyway.


    Zitat von jmateos

    Is there an official document available for the upgrade?

    Yep, you execute omv-release-upgrade. There are a couple of omv-extras plugins that should be removed because they haven't been ported or don't have upgrade paths.


    Zitat von jmateos

    Once upgraded , i could surely resize my 32 Tb partition ?

    Pretty sure but I don't have a 32tb ext4 partition to test on. If you used btrfs, xfs, or zfs, then I would say yes for sure.

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

    When i create a FS on OMV interface, i don't have the option of create differente Filesystems from a raid, the options are the device, label, and the FS type. And creates a FS with the entire size of the available RAID.


    So, the configuration that you suggest is? :


    Create a RAID6 FS with 8 x 7.TB disks on OMV.
    partition the MD device on several parts, with fdisk on debian console, format , and then make a mergerfs device with all of them.



    OR:


    create 2 or more RAID5/6 devices
    create a partition for each, format, and merge on the mergerfs partition


    On this option i may lose lot of space, 1 disk for raid5 and 2 for a raid6 device for each.





    Do you think on this case xfs has some advantages over ext4?





    Another question. I added the new disk to the existing RAID, but at the moment i cannot grow the FS. So, ¿It's possible to get off the RAID the disk?. Or the RAID will be on degraded status until i add the disk again?. Wil be this cause problems on the system?


    Thx

    • Offizieller Beitrag
    Zitat von jmateos

    Create a RAID6 FS with 8 x 7.TB disks on OMV.
    partition the MD device on several parts, with fdisk on debian console, format , and then make a mergerfs device with all of them.

    Don't do this. This might help against filesystem corruption but array rebuilds would still take forever.


    create 2 or more RAID5/6 devices

    Zitat von jmateos

    create a partition for each, format, and merge on the mergerfs partition


    On this option i may lose lot of space, 1 disk for raid5 and 2 for a raid6 device for each.

    If you create two 4 drive raid 5 arrays, you have the same loss of space as one 8 drive raid 6 array. If you do this, create a filesystem on each array and pool them with mergerfs.


    Zitat von jmateos


    Do you think on this case xfs has some advantages over ext4?

    xfs is always 64 bit and might work a bit better with large files but overall, I think they are about the same.


    Zitat von jmateos

    I added the new disk to the existing RAID, but at the moment i cannot grow the FS. So, ¿It's possible to get off the RAID the disk?. Or the RAID will be on degraded status until i add the disk again?. Wil be this cause problems on the system?

    If you didn't grow the filesystem, you should be able to remove the disk. Always makes me nervous though. What is the output of: cat /proc/mdstat

    omv 7.0.5-1 sandworm | 64 bit | 6.8 proxmox kernel

    plugins :: omvextrasorg 7.0 | kvm 7.0.13 | compose 7.1.6 | k8s 7.1.0-3 | cputemp 7.0.1 | mergerfs 7.0.4


    omv-extras.org plugins source code and issue tracker - github - changelogs


    Please try ctrl-shift-R and read this before posting a question.

    Please put your OMV system details in your signature.
    Please don't PM for support... Too many PMs!

Jetzt mitmachen!

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

Benutzerkonto erstellen Anmelden