NVME Crashing with "IO Error" (Dmesg); Causing data corruption

VIM4
1

Which system do you use? Android, Ubuntu, OOWOW or others?

Ubuntu

Which version of system do you use? Khadas official images, self built images, or others?

Official

Please describe your issue below:

I’m having IO errors with my NVME drive, using the new M2x board. Whatever’s causing the error leads to the NVME being unmounted, and then remounted as read-only (which is how I have configured it to mount if there’s an error, via fstab).

This NVME drive is being used as mass storage for docker containers, so having it go down at random is … not ideal. I’ve had to run fsck on it at least 3 times now, and lost a minecraft world to data corruption.

I haven’t figured out what triggers it, yet.

I’d appreciate any suggestions. I’m not sure how to diagnose this, so I’ve included the relevant dmesg output below.

Post a console log of your issue below:

~$ dmesg | grep -i nvme
[    0.698629] nvme nvme0: pci function 0000:01:00.0
[    0.698686] nvme 0000:01:00.0: enabling device (0000 -> 0002)
[    0.847912] nvme nvme0: Shutdown timeout set to 10 seconds
[    0.854462] nvme nvme0: 8/0/0 default/read/poll queues
[    0.867407]  nvme0n1: p1 p2
[    5.862197] EXT4-fs (nvme0n1p1): mounted filesystem with ordered data mode. Opts: errors=remount-ro
[    6.963007] EXT4-fs warning (device nvme0n1p2): ext4_clear_journal_err:5244: Filesystem error recorded from previous mount: IO failure
[    6.963599] EXT4-fs warning (device nvme0n1p2): ext4_clear_journal_err:5246: Marking fs in need of filesystem check.
[    6.983375] EXT4-fs (nvme0n1p2): warning: mounting fs with errors, running e2fsck is recommended
[    7.023987] EXT4-fs (nvme0n1p2): Errors on filesystem, clearing orphan list.
[    7.048644] EXT4-fs (nvme0n1p2): recovery complete
[    7.049162] EXT4-fs (nvme0n1p2): mounted filesystem with ordered data mode. Opts: errors=remount-ro
[  316.948911] EXT4-fs (nvme0n1p2): error count since last fsck: 2
[  316.948943] EXT4-fs (nvme0n1p2): initial error at time 1656687719: ext4_journal_check_start:61
[  316.948948] EXT4-fs (nvme0n1p2): last error at time 1656687719: ext4_journal_check_start:61
2

Hello @SinisterPisces

Are you test with latest 220620 imgae ?
What’s your SSD model?

1 Like
3

I … don’t think I have that image? How do I check?

The NVME is a Sabrent Rocket 4.0 1TB. It sits around 40 C.

1 Like
4

This is the same NVMe I’m using, and working out issues. It’s flawless on a StationPC M2 Rockchip ARM SBC. Both my M2 and VIM4 are formatted with btrfs.

I have stable boots on the VIM4, and also those that mount read-only, or spontaneously reboot.

The controller is identical to some known, good SSDs for VIM4. I’m optimistic about getting a stable and production capable build.

— Jeremiah

5

BTW: im not recommend btrfs to use anybody :wink: IMHO

1 Like
6

Just use OOWOW to install latest OS.

7

I’m a pretty old btrfs hand! You definitely have to know what you’re doing with it. It’s unforgiving of casual administration errors. Once I set it up, it works a lot like Apple’s new filesystem, and I use it on media workstations, for much the same ends.

8

okay , maybe if u know what u need from btrfs - its another story :wink:

1 Like
9

Haha, definitely seconding this!