Filesystem Mountpoint Missing

    • Offizieller Beitrag
    Zitat von claredad

    Yes, all three are the same. So you think overtime I should replace all three? Ill replace the first one now for sure.

    Following are the SMART Counts that should cause real concern


    SMART 5 – Reallocated_Sector_Count.

    SMART 187 – Reported_Uncorrectable_Errors.

    SMART 188 – Command_Timeout.

    SMART 197 – Current_Pending_Sector_Count.

    SMART 198 – Offline_Uncorrectable.


    SMART 199 - UltraDMA CRC errors

    (Usually hardware or cable related)

    _____________________________________________________________________


    Setting 199 aside, when any of the above begin to increment upward, around 5 to 10 or maybe a bit more, I'd be ordering another drive. The stat's for the drive you posted are not good. I wouldn't be surprised if you have a bit of data corruption already.


    If "ALL" of your drives are failing at once, in the same way, that's a statistical outlier. (I'd like to the stat's for all drives posted.) The odds of 3 drives failing in the same manner, at the same time, are pretty long (but not impossible). The CRC errors (SMART 199), especially if they're present on all 3 drives, has me wondering if these errors started after a power outage.

    Do you have your device on surge protection?
    _____________________________________________________________________


    geaves mentioned that these are enterprise class drives, but you said you didn't pay enterprise money for them. Refurb's or "white label" drives are not a bargain. Utilities exist to rewrite SMART stat's so it's impossible to know what you're getting. It's a "buyer beware" proposition.


    Users are better off with new, but slower, desktop drives or, as I like to put, "whatever is on sale".

    Thanks for taking the time to respond crash test.


    Yes, I went with refurbed enterprise drives, which in retrospect was a false economy. I have a WD RED now ordered to replace this one...will be here the start of January.


    I assume its safest to now power off the server and leave it off until I get the new HDD?


    Yes, we are prone to electricity failures here, and I think this is when my issues started. I have surge protection, but no UPS. Maybe that should be on my list on buys for the new year?


    Would you like me to run extended tests on the other two drives? Below is the basic info.


    SDE Drive: https://pastebin.com/uANmHhQm

    SDB Drive: https://pastebin.com/NArrGUwX

    • Offizieller Beitrag
    Zitat von claredad

    Would you like me to run extended tests on the other two drives? Below is the basic info.

    It wouldn't hurt anything to run extended tests, but those 2 drives appear to be OK.
    There are CRC's on one of them but that, in itself, is not a big deal. It could be a SATA cable. Try reseating the cable. (I would keep an eye on it. It might be an early symptom of a drive interface board failure.)

    BTW: mdadm RAID is smart enough to deal with mixed drives as in 2 drives at 7200rpm with 1 drive at 5400rpm. But the replacement will have to be as large or larger than the other 2.

    Scheduling SMART tests as geaves mentioned is a good idea. With automated SMART testing and server "Notifications" set up, the server will e-mail you when SMART errors increment.

    If you don't currently have it, you might want to give serious thought to backup. As you've found RAID is NOT backup.

    Thanks again for your help. I assume its prudent not to use the server until I get my replacement HDD, and get the RAID set up with 3 drives again?

    • Offizieller Beitrag
    Zitat von claredad

    Thanks again for your help. I assume its prudent not to use the server until I get my replacement HDD, and get the RAID set up with 3 drives again?

    That might be the prudent action. You could use the array but if there's a second drive issue, it's all over. Your call.

    Note that resilvering a 6TB drive is going to take a lot of time. During that time, the two remaining drives in the array are going to go through a torture test. This is what you should be worried about. (Remember, they may have more age on them than you know.) All you can do is hope a second drive doesn't fail, but I believe you'll have a decent chance of success.
    ___________________________________________________

    In the short term, you really need to set up notifications on your server. Check out this Guide, on page 35 and 36. The sooner you're aware of a drive issue, the better.

    I know it's expensive when talking about a 12TB array but, if you want to keep your data, you need to think about backup. Further, you might want to look into ZFS or, maybe, SNAPRAID for protection against silent corruption and bit-rot. In this regard mdadm RAID is doing nothing for you.

    Happy Christmas. Sorry about the long delay in responding... difficult Christmas with illness in the house. Not the virus though thankfully!


    My new 6TB drive arrived, and I want to fit it and use Snapraid. Reading the Snapraid documentation, it looks impressive, but starting with an existing RAID, I'm not sure where I'm supposed to start from my current situation. Do I repair the current RAID with the new disk first? Do I then just install Snapraid along with the current RAID?


    Sorry for what are probably stupid questions!

    • Offizieller Beitrag

    If you have all 10TB backed up, proceeding will be relatively easy but I would CHECK THE BACKUP (really check it) to be sure it's good.

    Note that SNAPRAID does not aggregate (or pool) drives like mdadm RAID does. You can use them individually, setting shares on each drive, or pool them together using Mergerfs. So, do you plan to use MergerFS to pool drives?

    Zitat von crashtest

    If you have all 10TB backed up, proceeding will be relatively easy but I would CHECK THE BACKUP (really check it) to be sure it's good.

    Note that SNAPRAID does not aggregate (or pool) drives like mdadm RAID does. You can use them individually, setting shares on each drive, or pool them together using Mergerfs. So, do you plan to use MergerFS to pool drives?

    Yes, I was planning to use MergerFS.


    I can visibly spot check files in the backup, but it's there any way to automatically check the backup to be sure it's good?

    • Offizieller Beitrag

    I haven't had the need to do this before, but something like diff might help in comparing the source to the backup. -> Compare


    When it comes to file corruption or bit-rot, it's likely the source is the same as the backup. The only way to be sure is to open data files in their app's.

    Zitat von crashtest

    I haven't had the need to do this before, but something like diff might help in comparing the source to the backup. -> Compare


    When it comes to file corruption or bit-rot, it's likely the source is the same as the backup. The only way to be sure is to open data files in their app's.

    Ok, thanks for all your help. Like geaves, you are a true credit to the OMV community the way you provide such great support.


    I'm planning to do a fresh install of OMV. Can I just remove the current RAID, then do the fresh install, setting up MergerFS and Snapraid as part of the new setup?

    Difficult getting time to do things here, but I have the fresh install now.


    When I did the fresh install the degraded RAID was still showing up, and no option to remove. I used the mdadm command you previously suggested, and that stopped the RAID. Today, Im going to try and set up snapraid and mergerfs.

    • Offizieller Beitrag
    Zitat von claredad

    I used the mdadm command you previously suggested, and that stopped the RAID

    I should have posted before you got that far with a clean install, you'll need to wipe the drives before proceeding any further and that includes the new one.

    • Offizieller Beitrag
    Zitat von claredad

    Today, Im going to try and set up snapraid and mergerfs.

    If you need a couple hints, take a look at this post. In that post, there's another link to the actual commands I used in Scheduled Jobs.


    While it's more advanced, you might look into some of the diff scripts. A "Diff script" looks at the number of changed files, since the last scrub, and will stop the scrub at a specified changed file threshold. (A scrub resets checksums and content files.) The idea is, if there's be a lot of corruption on a protected drive (many changed files) the scrub operation is halted so a drive recovery, if needed, is possible.


    (Edit: This -> post has the schedule I used, without a Diff script.)

    Zitat von geaves

    I've seen the thread, it's an NTFS drive, does have spaces in the Label, i.e. usb backup

    No, the label is just "Elements". Before I did the new install of OMV, this mounted without any issues.


    When its mounted on OMV (but I cant see it as an option in Shared folders), I can access the mount through Cloud Commander and move/copy files.

Jetzt mitmachen!

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

Benutzerkonto erstellen Anmelden