Crashing when transfering files

    • Offizieller Beitrag

    what chipset does inatek use

    Pass, but by the power of google I did find this and according to that, ->

    Zitat

    Inateck is using an ASMedia Technology ASM1153E bridge chipset, which is the latest (3rd) generation USB 3.0 to SATA II bridge chipset offered by the company and is more power efficient than the older ASM1053E bridge chipset that is widely used on existing docking stations.


    Since the mini PC has an M.2 slot, I thought about this

    Hm, again it will either work or it won't being an unbranded chinese manufacture

    Everything works correctly! It's definitely Docking

    Sorry I've not read the thread in it's entirety, TBH a docking station is not the best option, some users are using 2 and 4 bay icy box enclosures without any problems. I bought my dock station purely for it's cloning ability and I've upgraded a few Windows drives using it.

  • some users are using 2 and 4 bay icy box enclosures without any problems.

    Apparently Qoltec is doing something weird with the driver. I was obviously unlucky. If I had known, I would have taken something with an ASM chipset right away. I hope to be able to replace that Docking. My main goal was to avoid portable HDDs and use a better 3.5" one

    • Offizieller Beitrag

    The Qoltec, the UNITEK and my own inatek do appear to be similar both in their look and functionality, but I have never used mine as a permanently connected device

    This seems happen a lot - I mentioned the same thing earlier in this thread. I bought an Inateck that, over the space of a few years, seems to have transformed into a Sabrent model. I suspect the real difference to be nothing more than another vender putting their name on the case.

    ____________________________________________________________________________________

    I found this in local shops and it is available to me.

    This one is USB 3.2 which is faster than USB3.0 by about 15GB/sec. (-> Theoretical max is 20gb/sec for USB3.2 versus 5gb/sec for 3.0.)

    In the real world, with most spinning drives maxing out at 6gb/sec (that's a burst speed when reading from drive cache, for less than a second). Considering that the SATA interface is capped at a theoretical 6gb/sec, there probably won't be any noticeable difference between USB 3.2 and 3.0. With the typical consumer drive, after the cache is flushed, I/O goes way down to maybe 70MB/s. In the bottom line, with the spinning drive you have, you're unlikely to see any difference.

    _______________________________________________________________________________________

    The UNITEK Y-3026 has a second drive slot. If you need to expand at some point, that might be a really useful feature.

  • I returned the Qoltec 50316 and took the LC-35U3-C - USB-C Enclosure

    - Chipset ASM1153E


    It was available to me immediately, for everything else I would have to wait 15 days for delivery.


    I hope this solves my problems.

  • NOT! the problem was not solved.


    The same problem occurs again on both systems.


    On a mini PC and on an intel i3, 8gb ram system with an ethernet card.


    Regardless of the system, the problem is the same!

  • The speed drops in a few seconds from 100MB/s to some 5MB/s transfer

    After a few seconds, the system can crash.


    - I am connected to a 1Gbit switch, CAT5e cable.

    I reproduce the error every time by reading/writing a few GB constantly using SMB.


    Is SMB the one to bring down OMV?


    I'm starting to get crazy ideas because this is happening on 2 systems!

    • Offizieller Beitrag

    Is SMB the one to bring down OMV?

    Highly doubtful and I do mean highly doubtful. With 10's to 100,s of thousands of installations, if there was an SMB problem, this forum would be lighting up.

    Further, you can't simply ignore that memtest86 freezes on the minipc when you run it. That's another utility that has 100's of thousands of users. This is jumping around "hoping" that something is not wrong. I'm assuming that you've installed memtest86 on a thumbdrive. Does the same "freeze" happen on the i3 platform.
    ________________________________________________________________________

    Explain the following, leaving out no detail.

    The same problem occurs again on both systems.


    On a mini PC and on an intel i3, 8gb ram system with an ethernet card.

    I.E.

    - You're copying, over the network, between these systems, using an SMB network share as the source and the minipc as the destination. When a few GB of data, the attached USB drive disconnects. (Now, we know that the dock is not a likely cause of the problem.)
    - We also know that the freeze (or drive disconnect) happens when using a thumbdrive as a data source, when copying to the external drive. (Without the USB ethernet adaptor attached.)

    Does the same things (above) happen, in the same way, on the i3 platform?


    Lastly, the 4TB drive can not be summarily ruled out either. CRC errors can easily be generated by a hard drive's SATA interface card. It would be rare with a new drive but this issue is already outside the norm.

  • Does the same "freeze" happen on the i3 platform.

    yes.

    __________________________

    let me explain all again:


    I have 3 computers:

    -mini pc

    -desktop PC (i3, 8gb ram, ethernet card)

    -laptop for everyday use.


    I installed OMV on a mini pc and used usb HDD docking (jMicron) + 4TB WD purple + USB ethernet adapter.


    I connected to SMB from a laptop for everyday use.

    Read and wrote several GB of data.


    Laptop <--SMB--> mini pc


    After some time the disk unmount happened and the system log showed I/O error etc. (the system log is on my 1st post here)

    -everything blocked and I had to restart the mini pc.


    -I also tried WiFi on mini pc (no usb ethernet), the same thing happens.

    ___________________________


    After that I installed OMV on a desktop pc and used HDD docking (jMicron) + 4TB WD purple + USB ethernet adapter


    I connected to SMB from a laptop for everyday use.

    Read and wrote several GB of data.


    Laptop <--SMB--> Desktop PC


    After some time the disk unmount happened and the system log showed I/O error etc.


    -everything blocked and I had to restart Desktop PC

    ______________________________


    After that I tested OMV on a desktop pc and used HDD docking (jMicron) + 4TB WD purple + AND NO USB ethernet adapter BUT ETHERNET CARD OF THAT PC (I eliminated the USB ethernet adapter).


    I connected to SMB from a laptop for everyday use.

    Read and wrote several GB of data.


    Laptop <--SMB--> Desktop PC



    After some time the disk unmount happened and the system log showed I/O error etc.


    -everything blocked and I had to restart Desktop PC


    ______________________________


    I returned the Qoltec 50316 (jMicron) and took the LC-35U3-C - USB-C Enclosure

    - Chipset ASM1153E



    I connected the new HDD enclosure to the mini PC, made a quick wipe. After wiping I started to mount EXT4. It keeps getting stuck somewhere at "writing inode tables". Everytime at different progress.



    I gave up, then I connected the HDD and the new enclosure to a desktop PC (no ethernet usb adapter).


    Wipe disk, mount EXT4 went fine.


    I connected to SMB from a laptop for everyday use.


    Read and wrote several GB of data.


    Laptop <--SMB--> Desktop PC


    After some time the disk unmount happened and the system log showed I/O error etc.


    Nothing blocked this time. The disk was unmounted and remounted automatically, but it was no longer "sda" but "sdb". I couldn't access via SMB, I had to reset everything. IO processes are not stuck at 100% like before.


    ______________________________


    After that I did a fresh install of OMV on the mini PC.

    I installed OMV on a mini pc and used usb HDD enclosure (ASM1153E) + 4TB WD purple + USB ethernet adapter.


    Wipe disk, mount EXT4 went fine.


    I connected to SMB from a laptop for everyday use.


    Read and wrote several GB of data.


    Laptop <--SMB--> mini PC


    After some time the disk unmount happened and the system log showed I/O error etc.


    Nothing blocked this time. The disk was unmounted and remounted automatically, but it was no longer "sda" but "sda1". I couldn't access via SMB, I had to reset everything. IO processes are not stuck at 100% like before.


    ___________________________



    In between all that, I did a S.M.A.R.T disk test, which showed that there were no bad sectors or anything else.


    __________________________


    Now I'm doing a secure wipe because for some reason none of my PCs want to mount EXT4 anymore

    • Offizieller Beitrag

    1. What version of OMV are you using?
    2. How are you building OMV? Are you using the guide appropriate for your hardware which, if the OS is installed to a flash drive, includes testing boot media? (Hopefully, if you're booting from flash media, the media is new and of good quality. SanDisk, Samsung, etc., not a generic.) Weird, and I do mean truly weird, things happen when defective flash media is used for a boot drive.

    3. Did you do the "Thumbdrive data to the I3" copy test? (Not over the network.)
    4. If you're booting from a Thumbdrive; did you boot from the same thumbdrive, on the minipc AND the i3?
    5. The 4TB hard drive, while new, could be the issue. A defective hard drive, out of the box, happens. (Remember the CRC errors which may be at the root of this - we're talking about I/O errors. ) Do you have another hard drive to test with? Any kind or size of SATA drive would do.


    The disk was unmounted and remounted automatically, but it was no longer "sda" but "sda1".
    (And)
    The disk was unmounted and remounted automatically, but it was no longer "sda" but "sdb".

    This, believe it or not, is common. Linux assigns device names based on how they're presented by BIOS (UEFI) on a first come first served basis. This is why OMV disks are mounted by UUID under the /srv directory on the boot drive.

    As an example of a drive mount from one of my servers:
    /srv/dev-disk-by-uuid-a8d5de44-c9f1-421f-a4d0-be6e0cef08d3

    UUID's do not change.


    Now I'm doing a secure wipe because for some reason none of my PCs want to mount EXT4 anymore

    As the CRC's go up, I'm guessing. You know, if CRC's are still incrementing, with a new drive dock, that points to the 4TB drive.

  • 1. What version of OMV are you using?

    openmediavault_6.0.24-amd64 --> stable version


    2. How are you building OMV? Are you using the guide appropriate for your hardware which, if the OS is installed to a flash drive, includes testing boot media? (Hopefully, if you're booting from flash media, the media is new and of good quality. SanDisk, Samsung, etc., not a generic.) Weird, and I do mean truly weird, things happen when defective flash media is used for a boot drive.

    Yes.

    On the mini PC I installed OMV on its disk. I also tried installing and booting from Thumbdrive.

    No difference.

    On the i3 system, I again used a completely different Thumbdrive for booting.


    3. Did you do the "Thumbdrive data to the I3" copy test? (Not over the network.)

    Using rsync? I haven't tried it.


    4. If you're booting from a Thumbdrive; did you boot from the same thumbdrive, on the minipc AND the i3?

    On the mini PC I installed OMV on its disk. I also tried installing and booting from Thumbdrive.

    No difference.

    On the i3 system, I again used a completely different Thumbdrive for booting.


    5. The 4TB hard drive, while new, could be the issue. A defective hard drive, out of the box, happens. (Remember the CRC errors which may be at the root of this - we're talking about I/O errors. ) Do you have another hard drive to test with? Any kind or size of SATA drive would do.

    CRC is still at 30. Status has not changed.

  • After SMB stops and the disk is unmounted. I get this message in OMV.



  • I tried another HDD.

    Made a wipe and tried to mount.


    It doesn't want to mount again.


    Error:


  • Let's close this thread.


    Obviously, a lot of things happened to me that you need to pay attention to, and let this be a lesson to everyone.



    #1 jMicron HDD dock


    The main cause of the problem was the jMicron hdd dock, which used to disconnect itself in the middle of reading/writing and never connect again.


    Even when the PC was left on during the night and there was no reading/writing, I would find in the morning that the HDD dock had disconnected.


    S.M.A.R.T showed 30 CRC errors. With the new case ASM chipset there are no more CRC errors.


    -must use ASMedia chipset!



    #2 mini PC definitely failed


    Simply everything that could go wrong with a mini PC did. In the process of enabling OMV on a mini PC, the SSD memory failed, and the question is how many USB ports were "healthy".


    Currently the SSD shows 128MB, I can't access it or format it (and I don't want to bother) - it's going to the trash.


    #3 Desktop PC i3 system crashed


    Just crashed like the mini pc for a simple reason - OMV was installed on USB memory (flash)


    - it turned out that there is a problem when the system is booted from a USB flash memory (there was something wrong). I sacrificed the disk on the Desktop PC and installed OMV directly on the disk.


    Regardless of the fact that it was a completely different and new USB flash memory or old one.


    The USB flash did not show any errors, all tests of the bootable USB showed that it was correct.

    I just went round and round in circles.


    #4 ASM chipset is much more stable


    -ASM chipset currently puts the disk to sleep, when I start to write something, the disk spins again and everything works as it should. There is no strange unmount when it goes to rest.



    Conclusion:


    - a lot of things happened to me that led to problems. It turned out that I was spinning in a circle from which I could not get out.


    -Definitely boot the system from a disk that is connected to a SATA port or similar. I do not recommend USB boot at all!


    - jMicron is a problem because strange things happened even in a state where there was no reading/writing.


    - there is a chance that the external USB ethernet adapter made things even worse.



    THANKS to everyone for your help, I am slowly starting to "build" a different OMV hardware system where I will try to avoid such problems.

    • Offizieller Beitrag

    it turned out that there is a problem when the system is booted from a USB flash memory (there was something wrong)

    Can you give more details about this, please? I was surprised, I think it's the first time I read this.

  • Can you give more details about this, please? I was surprised, I think it's the first time I read this.

    I have no idea, I didn't investigate further.

    Ever since I installed OMV on the i3 PC's SSD drive, everything has been stable and working normally.



    Can you help me with another thing?


    I want to prevent the HDD from going into spin down --> using WD Purple designed for 24/7 spin.

    Start_Stop_Count -O--CK 100 100 000 - 174 --> is constantly increasing


    I use LC-35U3-C - USB-C Enclosure

    Chipset ASM1153E

    LC-35U3-C


    S.M.A.R.T is turned off, I set AMP to 255 - Disabled


    I tried [How-To] Spin-down hard drives with hd-idle and I set HD_IDLE_OPTS="-i 0 -l /var/log/hd-idle.log”


    use the parameter -i 0 to disable spin-down of not explicitly specified HDDs.


    The disk still goes into spin down

    It's probably the firmware?


    Can I set some coron or something to spin the disk 24/7?

    I think 24/7 is better than constant spin down and spin up

    • Offizieller Beitrag

    I have no idea, I didn't investigate further.

    Ever since I installed OMV on the i3 PC's SSD drive, everything has been stable and working normally.

    So I guess it must have been a specific problem with your hardware.

    I want to prevent the HDD from going into spin down --> using WD Purple designed for 24/7 spin.

    Start_Stop_Count -O--CK 100 100 000 - 174 --> is constantly increasing


    I use LC-35U3-C - USB-C Enclosure ...

    Do a search in the forum (better to search with google, the internal search engine does not have many hits...), this topic has already come up other times. I remember solutions such as a script or a scheduled task that from time to time "touches" the disk.

  • Do a search in the forum (better to search with google, the internal search engine does not have many hits...), this topic has already come up other times. I remember solutions such as a script or a scheduled task that from time to time "touches" the disk.

    I found this:By macom

    "One not so elegant solution would be to touch a file on the drive using a scheduled job every x minutes. WIth x smaller than the spin down time it should prevent the drive from spinning down."

    touch /srv/dev-disk-by-label-xxxx/stop_spin_down_file

    • Offizieller Beitrag

    I found this:By macom

    That's it, I was looking but I couldn't find it. That can be your solution.

    USB to SATA controllers are a nest of problems.

  • OK, this problem has never given me peace.


    I built a whole new system. Intel i3 processor, 12GB RAM, 1Gbps ethernet port, etc. I even bought a new WD Purple 4TB HDD and a new usb 3.0 docking station Chipset ASM1153E.


    I put everything together and said let's see if the same problem that bothered me before will happen again. I transferred 8GB via SMB and BUM! Exactly the same problem as before!


    I thought that the old mini PC was the problem, that the docking was the problem, and maybe even the HDD even though it was new!


    Everything blocked again, I/O went to 100% and SMB transfer speed dropped to 0 MB/s

    -identical problem as before even though the complete system is completely new? Yes! Absolutely yes!


    What is the explanation for this? Completely different PC, completely different docking, completely different HDD and the same error again?


    I don't understand why the external drive never works for me? There are a lot of users here for whom it works even on Rpi? - Crazy!


    The USB external drive is not usable at all! How can it happen that a completely identical thing happens on a completely different system and with completely different components? There is a deeper problem here, I believe.




  • Every SMB transfer causes the same error. Independent of hardware, independent of file size.


    The only thing that is different is that the HDD docking is connected to usb 3.0


    This is already the 3rd hardware on which the same thing is happening.

Jetzt mitmachen!

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