Scheduled Maintenance: We are aware of an issue with Google, AOL, and Yahoo services as email providers which are blocking new registrations. We are trying to fix the issue and we have several internal and external support tickets in process to resolve the issue. Please see: viewtopic.php?t=158230

 

 

 

Issue with a SSD model and Bullseye

Need help with peripherals or devices?
Post Reply
Message
Author
djiock
Posts: 3
Joined: 2022-07-25 13:23

Issue with a SSD model and Bullseye

#1 Post by djiock »

Hello !
I did some research on the internet but no clue, trying a post here as a last resort...

I work as a sysadmin for a cooperative where almost all our colleagues are using Debian on their computers (~120 PC). We're trying to keep up with Debian releases, and we're installing/upgrading to bullseye since a few months. We noticed after a while that a lot of computers had some serious Grub issues after a while, making boot impossible without running a fsck on boot partition, which got corrupted, and rebuilding Grub. And we have to do that for each computer every few weeks...

Trying to understand what was wrong, and realizing that not all models were affected we found out that the problem was with the SSD : Western Digital SSD WD Green 240 Go M.2 (WDS240G2G0B). When replacing it (with SATA SSD, that is maybe important), no more issue, even after several weeks.
I have some examples of computers that had no problem for a couple of years using Buster, then the issue above when reinstalled in Bullseye, and then fine when the WD SSD was replaced. The other models we have not using this WD SSD never had this issue.

Here are the most (I think) eloquent logs I could find :

Code: Select all

/var/log/syslog:Jun 27 11:00:01 ene-natl-pc-363 kernel: [ 4181.834080] EXT4-fs error (device sda2): htree_dirblock_to_tree:1054: inode #40962: block 331792: comm find: bad entry in directory: rec_len is smaller than minimal - offset=6144, inode=0, rec_len=0, size=1024 fake=0
/var/log/syslog:Jun 28 11:00:02 ene-natl-pc-363 kernel: [46751.776342] EXT4-fs error (device sda2): htree_dirblock_to_tree:1054: inode #40962: block 331792: comm find: bad entry in directory: rec_len is smaller than minimal - offset=6144, inode=0, rec_len=0, size=1024 fake=0
/var/log/syslog:Jun 29 11:00:01 ene-natl-pc-363 kernel: [79795.759952] EXT4-fs error (device sda2): htree_dirblock_to_tree:1054: inode #40962: block 331792: comm find: bad entry in directory: rec_len is smaller than minimal - offset=6144, inode=0, rec_len=0, size=1024 fake=0
/var/log/syslog:Jun 29 14:15:32 ene-natl-pc-363 kernel: [91526.696531] EXT4-fs (sda2): error count since last fsck: 3
/var/log/syslog:Jun 29 14:15:32 ene-natl-pc-363 kernel: [91526.696547] EXT4-fs (sda2): initial error at time 1656320401: htree_dirblock_to_tree:1054: inode 40962: block 331792
/var/log/syslog:Jun 29 14:15:32 ene-natl-pc-363 kernel: [91526.696561] EXT4-fs (sda2): last error at time 1656493201: htree_dirblock_to_tree:1054: inode 40962: block 331792
But I don't know how the issue is triggered...

If anyone has an idea I'd be happy to provide more information...

Thanks !

Edit : partitioning is done this way : EFI (FAT32) - boot (ext2) - "rescue" (ext2) - LVM2 LUKS encrypted, with root, var/log and home within

LE_746F6D617A7A69
Posts: 932
Joined: 2020-05-03 14:16
Has thanked: 7 times
Been thanked: 65 times

Re: Issue with a SSD model and Bullseye

#2 Post by LE_746F6D617A7A69 »

djiock wrote: 2022-07-25 14:02 Trying to understand what was wrong, and realizing that not all models were affected we found out that the problem was with the SSD : Western Digital SSD WD Green 240 Go M.2 (WDS240G2G0B)
Sounds similar to this one:
Problems with a Kingston A2000 1TB (NVMe SSD)
Bill Gates: "(...) In my case, I went to the garbage cans at the Computer Science Center and I fished out listings of their operating system."
The_full_story and Nothing_have_changed

djiock
Posts: 3
Joined: 2022-07-25 13:23

Re: Issue with a SSD model and Bullseye

#3 Post by djiock »

Hi ! Thanks for this quick reply, so we could try disabling APST ? Reading the thread I'm not sure it is the same issue, it doesn't look like I/O problems... but I'll look into it when I have a faulty PC at hand, I'm leaving for vacation for a few days.

LE_746F6D617A7A69
Posts: 932
Joined: 2020-05-03 14:16
Has thanked: 7 times
Been thanked: 65 times

Re: Issue with a SSD model and Bullseye

#4 Post by LE_746F6D617A7A69 »

Well, problems with damaged Ext4 entries/dir_entries are exactly matching those reported in that thread - that's why I've mentioned it.
Disabling APST is completely harmless - so You can try it just to confirm the problem or to reject one of possibilities ;)
Bill Gates: "(...) In my case, I went to the garbage cans at the Computer Science Center and I fished out listings of their operating system."
The_full_story and Nothing_have_changed

p.H
Global Moderator
Global Moderator
Posts: 3049
Joined: 2017-09-17 07:12
Has thanked: 5 times
Been thanked: 132 times

Re: Issue with a SSD model and Bullseye

#5 Post by p.H »

WDS240G2G0B has a SATA interface, not NVMe.

LE_746F6D617A7A69
Posts: 932
Joined: 2020-05-03 14:16
Has thanked: 7 times
Been thanked: 65 times

Re: Issue with a SSD model and Bullseye

#6 Post by LE_746F6D617A7A69 »

p.H wrote: 2022-07-26 20:55 WDS240G2G0B has a SATA interface, not NVMe.
According to this:
https://documents.westerndigital.com/co ... en-ssd.pdf

WDS240G2G0B means M.2 ...
Bill Gates: "(...) In my case, I went to the garbage cans at the Computer Science Center and I fished out listings of their operating system."
The_full_story and Nothing_have_changed

p.H
Global Moderator
Global Moderator
Posts: 3049
Joined: 2017-09-17 07:12
Has thanked: 5 times
Been thanked: 132 times

Re: Issue with a SSD model and Bullseye

#7 Post by p.H »

So what ? M.2 is a physical form factor, not an interface type.

LE_746F6D617A7A69
Posts: 932
Joined: 2020-05-03 14:16
Has thanked: 7 times
Been thanked: 65 times

Re: Issue with a SSD model and Bullseye

#8 Post by LE_746F6D617A7A69 »

p.H wrote: 2022-07-26 20:55 WDS240G2G0B has a SATA interface
Yes, You're right - somehow I misinterpreted the original post which is mentioning M.2 (and that PDF shows that too)
BTW, it's the 2nd case where I'm admitting that You're right ;)
Bill Gates: "(...) In my case, I went to the garbage cans at the Computer Science Center and I fished out listings of their operating system."
The_full_story and Nothing_have_changed

djiock
Posts: 3
Joined: 2022-07-25 13:23

Re: Issue with a SSD model and Bullseye

#9 Post by djiock »

p.H wrote: 2022-07-26 20:55 WDS240G2G0B has a SATA interface, not NVMe.
You're right ! I was wondering why I had /dev/sdax and not /dev/nvmexxx... So the issue is not linked with nvme, not the same case as the linked thread...

When replacing with another SSD I use a "classic" SATA interface which is also available in the laptop. So my issue could be with the M.2 controller or the SSD itself... Any idea what I could do to debug this ?

p.H
Global Moderator
Global Moderator
Posts: 3049
Joined: 2017-09-17 07:12
Has thanked: 5 times
Been thanked: 132 times

Re: Issue with a SSD model and Bullseye

#10 Post by p.H »

You could connect the SSD to the standard SATA port with a M.2-SATA adapter board and see what happens.

Are there kernel error messages related to the SSD (sda) but not to ext4 ?

Post Reply