server just unmounted my raid for no reason

1. offizieller Beitrag

    Hey guys I was just doing some cleaning of my raid then all of a sudden it fell off the network. I thought maybe it fell out of SMB but it turns out the WHOLE DRIVE WAS UNMOUNTED. I am not sure why this happended.


    Where can I go in the OS to see if there were errors or something on why it is doing this?

    • Offizieller Beitrag
    Zitat von digitalbots

    oddly enough I don't have a /dev/sde

    ?( this is from the text file you uploaded

    Code
    Mar 15 11:38:01 omv5 kernel: [ 5469.935856] EXT4-fs error (device sde1): ext4_journal_check_start:83: Detected aborted journal
Mar 15 11:38:01 omv5 kernel: [ 5469.938427] Buffer I/O error on dev sde1, logical block 2264925223, lost async page write
Mar 15 11:38:01 omv5 kernel: [ 5469.940950] EXT4-fs warning (device sde1): htree_dirblock_to_tree:1004: inode #141983844: lblock 0: comm Plex Media Serv: error -5 reading directory block
Mar 15 11:38:01 omv5 kernel: [ 5469.942523] Buffer I/O error on dev sde1, logical block 2275934307, lost async page write
Mar 15 11:38:01 omv5 kernel: [ 5469.945520] EXT4-fs warning (device sde1): htree_dirblock_to_tree:1004: inode #141951088: lblock 0: comm Plex Media Serv: error -5 reading directory block

    that's just the first few lines, then there's this

    then there's this

    Code
    Mar 15 11:38:01 omv5 systemd[1]: srv-dev\x2ddisk\x2dby\x2dlabel\x2dYottamaster.mount: Unit is bound to inactive unit dev-disk-by\x2dlabel-Yottamaster.device. Stopping, too.

    the above is saying that Yottamaster is inactive, but there appears to be an issue with one of the drives, so the raid has become inactive.

    Yottamaster has a hardware array built into it. So there are 5 drives attached to the yottamaster but when I plug it in my machine only sees it as /dev/sdk


    I can disable the hardware array then all 5 drives will show up in OMV and try a software array doing that would give me the ability to see each drive.


    here is my output from the df command


    • Offizieller Beitrag
    Zitat von digitalbots

    I can disable the hardware array then all 5 drives will show up in OMV and try a software array doing that would give me the ability to see each drive.

    I'm only guessing here, but the error seems to suggest that is exactly what happened, but OMV is not going to know why, it just knows it's inactive, when that occurred it appears it was capable to see each drive and spotted an error on one of them.


    But if you are running a hardware raid then then that should give you more information, and it should either repair itself or warn that there is a drive failing.


    Just found their website, is this one of those usb multi drive bays where you can set the raid type or use as single drives.


    If this has been running for some time you can't just switch over it's functionality without losing data.

    • Offizieller Beitrag
    Zitat von digitalbots

    its new array and brand new drives. i can dump without crying lol

    ^^ these things are fine if they are on windows or a mac, but on linux all you have is the unit itself which will just flash a light if there is a problem with a drive, but windows and macs they have software to help diagnosing a problem.


    This would be better being run as JBOD, then create either a software raid or use mergerfs + snapraid in omv's gui, but whatever you decide you will have wipe all the drives under storage -> disks before doing anything.

    • Offizieller Beitrag
    Zitat von digitalbots

    i have a 4 disk drive raid working via USB

    ?( did you create a software raid through omv


    Zitat von KM0201

    Must just be when it's USB drives in individual enclosures then

    AFAIK this should not make any difference and it wasn't until I just logged in that what I stated above would have to be performed from the CLI.

    ______________________________________________________________________________________


    Going back to #1 and the subsequent syslog I would surmise that this was caused by a power outage or power surge, causing the Raid setup via the USB box to go inactive. This meant that the box was seen as individual drives when omv came back up.


    Assuming the above is correct then a UPS would be a solution to hopefully prevent it from happening again.


    These units work on a sata to usb bridge backplane similar to the radxa sata hat, the only difference being is that the Raid functionality is controlled via the box, just like a hardware raid controller. This is the downside for Linux users as there is no GUI for that Raid setup, you're reliant on the unit giving you a constant 'pulsing light' to warn you of an error with a drive. Unlike Windows and Mac users where the manufacturers software GUI is available to install.

    • Offizieller Beitrag
    Zitat von digitalbots

    correct.

    Interesting

    Zitat von KM0201

    Must just be when it's USB drives in individual enclosures then

    Learn something new everyday, these types of units must be using a sata to usb bridge similar to the radxa

    Same happening to me, two 500GB disks connect on usb hub with external power supply are unmounted from the system after a period of inactivity or just after some time. I have to manually mount them to be able to access again.

Jetzt mitmachen!

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

Benutzerkonto erstellen Anmelden