Posts by FabrizioMaurizio

    Well, all flash memory will eventually wear out. If you put stuff on it resulting in regular writes and especially with a high 'Write Amplification' then it will wear out faster. The flashmemory plugin won't help in these cases and most probably you're not able to replace a worn out eMMC on the PCB.

    The eMMC is not soldered to the board so replacing it should not be a problem. The thing is: it feels kind a waste to use a 32gb emmc just for the operating system. A user in this forum suggested to proceed in this manner to move the docker folder:


    So my solution is :


    1) don't touch the setting in Docker tab
    2) Move /var/lib/docker to external drive (ex. /sharedfolders/data/docker)
    3) sym-lync /sharedfolders/data/docker to /var/lib/docker

    I did the measurements under load and the minimum was 4,85v. I also measured the other 2 adapters that I own (the one that was on the wd elements external drive (asm1051W) and the usb to dual sata adapter that I bought from ebay (jms561u)). Both these adapters work in different way from the one with the jms578. When I attach the usb to the sbc (only the usb, not the 12v power brick) the jms578 gets 4,97v at the sata connector, the other two adapters get 0v so I'm guessing they get the 5v by converting the 12v they get from the power brick.


    Anyway, these are the results (measurements taken at the sata connector with the drive under load):


    asm1051w - 4,95v
    jms561u - 5,05v-5,10v
    jms578 - 4,85v


    I'm now using the dual sata adapter (jms561u) as it should not be affected by undervolting issues.
    Questions:


    1. Is it possible to fix the jms578 adapter's undervolting issues?
    2. Is it possible that the freezes were caused by transmission since I changed the default values (max downloads: from 5 to 20, max peers per torrent:100, max peers overall:500)?
    3. Should I just ignore this error ERROR Transfer event for disabled endpoint or incorrect stream ring?

    You have a 0.2V drop between board's PSU and the adapter and another 0.15V between the adapter's USB side and the SATA power connector. Without any load if I understood correctly.

    that's correct. How do I apply a heavy load to the drive. I tested while muxing a video on the drive and the voltage dropped to 4,90v with the lowest peak at 4,85v at sata connector. Also, since the psu already delivers over 5v and the usb cable is pretty short, does this mean there is nothing I can do to fix this?

    I teared apart the adapter and did some measurements. I get 5,13v where the usb wires are soldered to the board and 4,97v to the sata connector. Also I get 5,35v from the rock64's power brick. Measurements were done with no drive attached. Also I can confirm it's a jms578. The usb cable is 30cm long, AWG is not written on the cable.

    The disk "disappearing" was my fault because I accidentally bumped into the cable and it got disconnected while downloading.


    I got the ERROR Transfer event for disabled endpoint or incorrect stream ring as a result of underpowering:

    I'm using a dedicated 4A power brick (that should be more than enough power) for the hard drive and the original pine64 3A power brick for the rock64. I get that error every time I boot the system up (but not with uasp blacklisted).
    The issue I'm having is that the nas becomes unreachable when the hard drive is under heavy load (sometimes i mux videos from and into the drive while downloading multiple files at the same time) and I have to force a reboot. What could cause this?



    Are the adapters really JMS578 (check lsusb output)?


    They should be the real deal.

    Code
    root@rock64:~# lsusb
    Bus 005 Device 002: ID 152d:0578 JMicron Technology Corp. / JMicron USA Technology Corp.
    Bus 005 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
    Bus 004 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
    Bus 003 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub
    Bus 002 Device 002: ID 152d:0578 JMicron Technology Corp. / JMicron USA Technology Corp.
    Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
    Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

    nevermind, I just did a fresh install. All seams to be working for now.


    edit.


    It looks like it wasn't working after all. It was better than before but not fully fixed. I'm trying a new thing now. I found this post where other users were having the same issue and I checked the system log and I had the same error they had.

    Code
    Jun 10 00:23:21 rock64 kernel: [ 50.374370] xhci-hcd xhci-hcd.9.auto: ERROR Transfer event for disabled endpoint or incorrect stream ring
    Jun 10 00:23:21 rock64 kernel: [ 50.374386] xhci-hcd xhci-hcd.9.auto: @00000000f2cc24b0 00000000 00000000 1b000000 01078001

    So, as they suggested I blacklisted uasp and the error disappeared and so far I haven't had any freezes. Will report back if this changes.


    What I fail to understand is why is this happening. I bought this adapter (3 of them actually) because it has the jms578 which is known to be a good chip. I flashed the latest firmware from odroid and all seemed to be working fine but apparently it wasn't. I also have a usb to dual sata adapter with the jms561u that someone reported working with the rock64. Well it does work but I get the same error as above unless i blacklist uasp (also hdparm can't recognize if a disk is active or in standby).


    Other than the chip and the firmware, what else should one look for when buying this things? @tkaiser @Adoby

    Looks like I have another issue to deal with :( . The nas becomes unreachable some minutes after it has booted. I cannot access the gui nor ssh and from the router's gui it appears like it's disconnected. I though this was caused by the low free space but I guess not. The only way to reach it again is to force a reboot by holding the power button on the board, but shortly after it reboots it becomes unreachable again.

    To allow you to use the last partition for data storage as explained at the download page: https://sourceforge.net/projec…ngle%20Board%20Computers/
    An OMV install is pretty fine with less than 4GB so those 8GB we limit the OS partition too are some sort of safety headroom. If you put your dockers on the OS drive on a SD card be prepared that the card might die pretty soon (flashmemory plugin won't help here at all).

    Not sure I understand how this should be setup. Should I format the remaining emmc storage with mkfs.btrfs /dev/mmcblk0p3 and then install the docker containers into that partition or should I use a different sd card completely (I have an empty sd card slot)?

    I accidentally bumped into a cable the other day and a drive got disconnected, maybe that's how that stuff ended in there. I also experimented with different sata to usb adapters so that also might have something to do with it.

    I booted with my drives disconnected.



    Looks like there is some stuff in sharedfolders. Can I simply delete it? I checked and it's old stuff I don't need.

    Hi, I have a rock64 with 4gb ram and 32gb emmc with OMV 4. I noticed one of the system partition is almost full, namely /dev/mmcblk0p2. I only have 1 docker container running (I had 2 but I uninstalled one to free up some space) and I already cleared system logs.



    Help please :) . Also, I have a 32gb emmc, why is my system partition only 7.4gb?