Jump to content

Odroid HC1 SATA disk switches between sda and sdb


xorinox

Recommended Posts

Hi, I have 3 Odroid HC1. All have a 2TB SATA disk attached. I am using Armbian_5.35_Odroidxu4_Ubuntu_xenial_next_4.9.61.img, and the armbian-config to copy the rootfs to the SATA disk. I have used ext4 and btrfs. Actually used btrfs from the start, but I got some weird error after running the HC1 for a couple of days, so that my open command, would return -bash: /sbin/blkid: Input/output error on any input. This is just an example. Didn't really know what was going on, I rebooted, and then could not login through ssh anymore. I assume rootfs gone, and reinstalled the HC1 again, copied over rootfs with armbian-config using btrfs.

 

Few days later I had the same issue again, so reinstalled again but changed to ext4 (to see if this makes a difference). Now I have that same issue also with ext4 as the rootfs. So it's unrelated to ext4/btrfs I assume. I noticed that the SATA disk from originally sda changed to sdb. Have a look at lsblk below. I restarted the node and it came back working, but now using sda again. You also find the dmesg outpout from when the node wasn't working properly anymore.

 

I think the trouble started at line: [ 7432.871964] usb 4-1: USB disconnect, device number 2                (maybe these disks go sleep)?   I am not running any big computations on them. The power supplies are 5V/4A

 

I have 2 HC1 with a 2TB Seagate Barracuda and the 3rd HC1 with a 2TB Seagate Firecuda. Both HC1 with each the Barracuda had that issue so far. One with ext4, one with btrfs. The disks are new. The setup with the Firecuda didn't break yet.

 

I have tried to read SMART but didn't get that working: smartctl -a -d usbjmicron /dev/sda

 

Any help much appreciated.

 

 

root@hc1-node-1:~# lsblk

NAME        MAJ:MIN RM   SIZE RO TYPE MOUNTPOINT

zram7       253:7    0 124.6M  0 disk [SWAP]

zram5       253:5    0 124.6M  0 disk [SWAP]

sdb           8:16   0   1.8T  0 disk

└─sdb1        8:17   0   1.8T  0 part /

zram3       253:3    0 124.6M  0 disk [SWAP]

zram1       253:1    0 124.6M  0 disk [SWAP]

mmcblk1     179:0    0  14.9G  0 disk

└─mmcblk1p1 179:1    0  14.7G  0 part /media/mmcboot

zram6       253:6    0 124.6M  0 disk [SWAP]

zram4       253:4    0 124.6M  0 disk [SWAP]

zram2       253:2    0 124.6M  0 disk [SWAP]

zram0       253:0    0 124.6M  0 disk [SWAP]

root@hc1-node-1:~# blkid

-bash: /sbin/blkid: Input/output error

root@hc1-node-1:~#

 

 

 

 

dmesg.txt

Link to comment
Share on other sites

11 hours ago, xorinox said:

usb 4-1: USB disconnect, device number 2                (maybe these disks go sleep)?   I am not running any big computations on them. The power supplies are 5V/4A

 

Usually a sign of either USB cable/receptacle crappiness (impossible on HC1) or undervoltage. You should check how much voltage your PSUs do really provide under load (Ohm's law). More info: https://forum.armbian.com/topic/3953-preview-generate-omv-images-for-sbc-with-armbian/?do=findComment&comment=32340

 

Link to comment
Share on other sites

Ok, thanks. I will find out if my psu are the reason. I don't have the tools to measure my PSU's performance under load. Yet need to read through the add'l information.

 

Meanwhile, I got an LRS-150F-5 and changed over 2 of 3 HC1. It provides up to 22A. As you can see the cables are thick copper cables too.img_20171228_185620_1024.thumb.jpg.8ba5de2b0eed22bb4b5a3ec1e9f111da.jpg

Link to comment
Share on other sites

2 hours ago, xorinox said:

Meanwhile, I got an LRS-150F-5 and changed over 2 of 3 HC1. It provides up to 22A

 

For 2 HC1 in a pure NAS scenario 6A are sufficient (keep this in mind when you measure consumption at the wall since your oversized PSU is highly inefficient in this low load range and wastes a lot of energy).

 

I write it again one last time: the problem is called underVOLTAGE (and not amperage too low). Meanwell PSUs have adjustable voltage so you need to ensure that your PSU is providing 5.2V and not 4.7V since otherwise you most probably run into the same issues again.

 

Edit: According to the LRS-150F-5 datasheet voltage is adjustable between 4.5V and 5.5V so without a multimeter it's not possible to make any reasonable use of this power supply (since if it's set to much lower than 5V undervoltage at the disk/controller might happen again)

Link to comment
Share on other sites

My understanding is, when the connected device tries drawing more amps, than the PSU under specified voltage can provide, the voltage collapses. My PSU from Amazon (https://www.amazon.com/gp/product/B01K0608A0) have 20W, if the HC1 under some circumstances draws more than 4A the volts have to go down, since its max is 20W (thinking about the power law).

 

I will buy a multi-meter and do the measurements suggested. The one HC1 connected the Meanwell PSU didn't show the issue yet. Do know, I am not suggesting this 150W PSU as a final solution.

Link to comment
Share on other sites

On 16.1.2018 at 4:48 PM, xorinox said:

My PSU from Amazon (https://www.amazon.com/gp/product/B01K0608A0) have 20W,

 

23% 'one star rating' at least for me is a very clear 'never ever buy such crap' indicator. 2nd 1 star review names the problem already:

Quote

No good for USB power. The output voltage drops 17.3%, from 5.2V to 4.3V, from no load (0A) to full load (4A). Way, way out of spec for USB power. (At least the one that was shipped to me does so).

 

As very often with cheap PSUs you can not trust in what's printed on them. 5V/4A would mean 20W if you multiply both numbers. But crappy PSUs either provide 5V (no load) or 4A but not both at the same time (voltage drop under load). And the same phenomenon can be observed with cables between PSU and board if wires are too tiny. Incoming voltage at the board dropping due to cable/contact resistance too high. And this happens only with some load generated (Ohm's law).

 

Can a moderator now please move this whole thread to the subforum where it belongs too? Thanks!

Link to comment
Share on other sites

The issue with sda vs sdb didn't come back since I am using the oversized power supply. I got a multi-meter and measured the voltage under normal conditions, when the HC1 is loaded with compiling code and used one of my utilities to create lots of files keeping the disk busy. The voltage stayed in between 5.19 and 5.17.

 

I noticed that my switch suddenly indicated that the Ethernet connection from it to HC1 is only of 100Mbit (orange light). I unplugged the cable and it changed back to 1Gbit. A little after it happened again. I am actually running Samba on the HC1 and indeed the connection seemed capped at 100MBit e.g. copying files from my Win machine didn't exceed 10Mbyes/s. After unplug and plug again it was back to 1GE and 50-70Mbytes/s transfer speed (from Win to it). I continue watching if I see the 1GE downgrade to fast Ethernet again.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
×
×
  • Create New...

Important Information

Terms of Use - Privacy Policy - Guidelines