USB errors when using USB 3.0 external disk and SMART

  • Hi folks,


    I'm using my OpenMediaVault server just as a repository for backups of my other Linux stuff using RSYNC. As a result, I used an Intel NUC Celeron 3050 as the vault server with an external USB 3.0 8TB drive enclosure.


    If I ever attempt to get S.M.A.R.T data from this disk, I get USB interface errors and USB resets. But in general, USB seems flaky and along with the bus resets I occasionally see "cmd cmplt err -2" from the UAS driver. I seem to have reduced the frequency of those by disabling USB legacy in the BIOS.


    I googled the heck out of this and can't see any specific reasons why I can't use USB for storage.



  • Which kernel version and which USB3 device (uname and lsusb)?
    Rsync error on external usb disk

    Linux vault 4.19.0-0.bpo.4-amd64 #1 SMP Debian 4.19.28-2~bpo9+1 (2019-03-27) x86_64 GNU/Linux


    lusb returns the following for the external disk:

  • Looks similar to the 'Seagate USB3 disks with Linux' issue (ATA command pass through) as such you could try to UAS blacklist the product and device ID: options usb-storage quirks=059f:106f:u as outlined in the link above.


    Please report back whether that helps.

    Tried that already. Did not make a difference to this issue.


    If I stay away from the SMART GUI, I don't get any USB errors. But if I go into SMART and click on "Devices" (even with SMART disabled) I get the resets.

  • But if I go into SMART and click on "Devices" (even with SMART disabled) I get the resets

    Should be reproducable then running one rsync task and smartctl -x for the disk device in question in parallel (behavior should be the same with or without UAS, just error messages will differ)?


    Most probably one of the many downsides of USB storage with Linux. IIRC I had similar problems with my JMS578 based USB3 enclosures but a firmware upgrade fixed it.

  • I gave it a quick try on a RockPro64 with an USB3 attached SSD in a VIA715 based enclosure:



    The command OMV is using when entering the device page is smartctl -x $device -d sat. I then created 2 GB of data consisting of +5000 files, rsynced them and constantly ran time smartctl -x /dev/sda -d sat. While execution of the smartctl command was drastically delayed (23 seconds was the longest), no issues occurred:


    Oops, seems the connection has only been established with USB2/Hi-Speed (Port 1: Dev 2, If 0, Class=Mass Storage, Driver=uas, 480M). So I need to retest later when I can inspect the device physically...

  • JMS578

    Hi guys, I have a more or less basic (maybe dumb) question: How do I know, which chipset ist used by an external USB3 enclosure? I found this information only in rare cases in the product discription.
    Sometimes it is mentioned, sometimes not. Sometimes the chipset has changed during production time, but the product has the same article number.


    Is it really the case, to figure it out by trial and error, or is there any other reasonable resource, which such kind of information?

    OMV 3.0.100 (Gray style)

    ASRock Rack C2550D4I C0-stepping - 16GB ECC - 6x WD RED 3TB (ZFS 2x3 Striped RaidZ1) - Fractal Design Node 304 -

    3x WD80EMAZ Snapraid / MergerFS-pool via eSATA - 4-Bay ICYCube MB561U3S-4S with fan-mod

  • How do I know, which chipset ist used by an external USB3 enclosure?

    You need to check vendor and product ID (lsusb on Linux, system_profiler SPUSBDataType or looking into "System Information.app" on macOS, most probably Device Manager on Windows). And this is a post purchase step unfortunately :)


    And the problem remains that certain renowned USB3 disk manufacturers like Seagate, WD or LaCie here use common chipsets from ASMedia, JMicron or VIA but flash their own branded firmware which changes the device's behavior. Seagate for example uses/used the ASM1153 bridge chip but all their disks have an own firmware not reporting ASMedia's vendor ID 174c but Seagates and reporting individual product IDs. Also behavior is different based on the firmware.


    My personal take on this is to buy by chipset. Only if the chipset is mentioned somewhere I'll order an enclosure. Side effect: no popular USB3 drives for me since you never know what's inside and there's little chance to get potential firmware issues fixed.

  • Should be reproducable then running one rsync task and smartctl -x for the disk device in question in parallel (behavior should be the same with or without UAS, just error messages will differ)?
    Most probably one of the many downsides of USB storage with Linux. IIRC I had similar problems with my JMS578 based USB3 enclosures but a firmware upgrade fixed it.

    OK I'll do that with both UAS and the quirk and see what the behaviour is. I'll also see if there's a firmware update for this enclosure.

  • Looks similar to the 'Seagate USB3 disks with Linux' issue (ATA command pass through) as such you could try to UAS blacklist the product and device ID: options usb-storage quirks=059f:106f:u as outlined in the link above.


    Please report back whether that helps.

    I had the same issue with an external selfpowered HD box (FANTEC QB-X2US3R ), I have a couple of 3T HDDs on it.


    And it crashes everytime.


    Tryed UAS blacklisting but with no result. I have also SMART disabled.


    Any idea how to proceed? (the same FANTEC thing goes fine on the USB2.0 port but painfully slow...).


    Best regards.

  • I had the same issue with an external selfpowered HD box (FANTEC QB-X2US3R ), I have a couple of 3T HDDs on it

    Are you still on an ODROID XU4? If so I would suspect you're victim of USB3-A receptacle crappiness (the XU4 is famous for). Also please keep in mind that the 2 USB3 ports on the XU4 are behind an internal USB hub which is something you should try to avoid.

Jetzt mitmachen!

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