Ext USB disk plugged out/in repeadtely

[mdi]

Hi,
I recently bought a new USB 2 docking station to replace the USB 3 i had, because the USB 3 one didn't work well with the NwSp2 v17.0.
The new model is Ronsen 895U3SC, it can host 2 sata disks and i have plugged in 2 3.5" disks of 2Tb each, one ext4 partition per disk.

What happens is that the disk 2 has been permanently unmounted, while disk 1 seems to work (i backup the NAS shares on it), but i receive "EXT USB disk plugged out" and "EXT USB disk plugged in" mails at least once a day, if not more often.

The disk is plugged out and back in in the same minute. It happens at different times during the day, so I don't know what causes this.

What/how should I investigate?

Thanks a lot.

[Jocko]

Hi mdi,

What happens is that the disk 2 has been permanently unmounted, while disk 1 seems to work (i backup the NAS shares on it), but i receive "EXT USB disk plugged out" and "EXT USB disk plugged in" mails at least once a day, if not more often.

That would mean the kernel does not support well your USB device.
So maybe there is more information in dmesg output.

Did you install the last rollup patch (version 08 06/12/18) ?

You may also try to update usb id database :

Code: Select all

/usr/sbin/update-usbids.sh

[fvdw]

To add;
In case it connects and is recognized and later it spontanuos disconnects it is not a software issue to my opinion, it could point more likely to a bad connection, either a broken cable or connector with bad contact. Try another cable if it can be exchanged. Also power supply of the usb box could be unstable.

[mdi]

I will verify it. Dmesg reports some filesystem issues, but inalso reformatted the partition...

Is this normal?:

root@nas:/ # lsusb
unable to initialize libusb: -99

[Jocko]

Hi mdi

lsusb is an out dated feature (and not safe). So it is disabled.

To get a similar output, use

Code: Select all

/usr/bin/usb-devices

But please to post your dmesg

[mdi]

Hi Jocko, i attached my dmesg.

I've temporarily lost the power and this is the dmesg after the black out.

Hope it reveals something.

In the meantime I opened the nas case and tried to saueeze a bit the usb connector which seemed slightly a bit loose.

Now the cord seems a bit more stable once connected. I hope the issue was only that.

Thanks,
MDI

[Jocko]

Ok mdi,

I did not find anything serious in your dmesg output.

But one thing may be easily fixed

Code: Select all

[  400.111366] EXT4-fs (sdc1): error count since last fsck: 339
[  400.117020] EXT4-fs (sdc1): initial error at time 1543990965: ext4_lookup:1602: inode 2
[  400.125025] EXT4-fs (sdc1): last error at time 1544328007: htree_dirblock_to_tree:987: inode 115867649: block 463478816

So you need to unmount first sdc1 :

Code: Select all

/etc/api/nasapi.php mrvlUdevdDetached sdc1

and run

Code: Select all

e2fsck -y /dev/sdc1

and try again to mount it

Code: Select all

/etc/api/nasapi.php mrvlUdevdAttached sdc1

Then see if you have still this warning in dmesg output

About your 2d disk, this one is rightly detected. What file system is set on sdd1? Please to post

Code: Select all

file -bs /dev/sdd1
fdisk -l /dev/sdd

.
We go to try to mount it manually:

Code: Select all

echo >/var/log/messages
debugon
/etc/api/nasapi.php mrvlUdevdAttached sdd1
debugoff

So post the output of these commands and also the content of /var/log/messages

[mdi]

Hi Jocko,
in fact now also the 2nd disk mounts properly. I yesterday run a fsck on it and it fixed the mount issue.

After doing it also on disk 1, dmesg is better:

[ 34.118201] EXT4-fs (sdd1): recovery complete
[ 34.123426] EXT4-fs (sdd1): mounted filesystem
with ordered data mode. Opts: user_xattr,acl,nod
elalloc
[ 36.294753] EXT4-fs (sdc1): recovery complete
[ 36.300051] EXT4-fs (sdc1): mounted filesystem
with ordered data mode. Opts: user_xattr,acl,nod
elalloc

So thanks a lot guys!
If i have more troubles, i will write again.
I'll be monitoring the docking station against unwanted un/mounts in the next days...

Thank you!

Here is the output you asked for:

root@nas:/ # file -bs /dev/sdd1
Linux rev 1.0 ext4 filesystem data, UUID=7ad7639b
-a6c8-45a8-93ed-055bd14eed5a, volume name "data2"
(needs journal recovery) (extents) (large files)
(huge files)
root@nas:/ # fdisk -l /dev/sdd

Disk /dev/sdd: 2000.3 GB, 2000398934016 bytes
256 heads, 63 sectors/track, 242251 cylinders
Units = cylinders of 16128 * 512 = 8257536 bytes

Device Boot Start End Blocks
Id System
/dev/sdd1 1 242252 1953514583
+ ee EFI GPT

[mdi]

Hi Jocko, after a couple of hours i got again a disconnection of the USB disks.
I attach the dmesg.

I suspect a disk fault, but it seem strange as the disk are almost never touched physically, so a hardware failure is improbable. But you never know :-)

[Jocko]

Hi mdi

We made some progresses.
It seems to be an issue with ext4 file system : The kernel fails to update the journal.

So can you post please :

Code: Select all

tune2fs -l /dev/sdf1
tune2fs -l /dev/sde1

(of course need to update with the last device kernel name if you got other events).

How do you format these disks : manually or use the firmware ?

other point: did you enable the usb spin down guard ? (if yes , you should try without enabling it)

And last ask, what is the current kernel version ?

Code: Select all

uname -a