[solved] proxmox 9 system disk crash, web gui not responding after 10 min with kernel above 6.8.12-13-pve

[denut]

hello


i use proxmox on my server since 8 version

motherboard : x399 designare ex

processor :Threadripper 1950x

boot disk on sata controler



all was working perferct

updating to proxmox 9

working again



updating via all kernel to 6.8.12-13

working again



sinse kernel 6.14.xx-x

the system crash after 5 or 45 min but never workin more than 1 hour

vm are on but web interface unrushable or rushable but console shell or vm shutdown/reboot not working





in the system log i have :

Oct 16 06:55:46 pve-tr4 kernel: ata1.00: failed command: READ FPDMA QUEUED
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: cmd 60/08:10:08:32:64/00:00:0c:00:00/40 tag 2 ncq dma 4096 in
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: status: { DRDY }
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: failed command: WRITE FPDMA QUEUED
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: cmd 61/08:48:08:10:5e/00:00:01:00:00/40 tag 9 ncq dma 4096 out
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)


and :



Oct 16 06:55:46 pve-tr4 kernel: ata1.00: failed command: WRITE FPDMA QUEUED
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: cmd 61/08:50:28:30:25/00:00:01:00:00/40 tag 10 ncq dma 4096 out
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: status: { DRDY }


and :



Oct 16 06:55:46 pve-tr4 kernel: ata1: hard resetting link
Oct 16 06:55:46 pve-tr4 kernel: ata1: SATA link down (SStatus 0 SControl 300)
Oct 16 06:55:46 pve-tr4 kernel: ata1: hard resetting link
Oct 16 06:55:46 pve-tr4 kernel: ata1: SATA link down (SStatus 0 SControl 300)
Oct 16 06:55:46 pve-tr4 kernel: ata1: limiting SATA link speed to <unknown>
Oct 16 06:55:46 pve-tr4 kernel: ata1: hard resetting link
Oct 16 06:55:46 pve-tr4 kernel: ata1: SATA link down (SStatus 0 SControl 3F0)
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: disable device




Oct 16 06:55:46 pve-tr4 kernel: Buffer I/O error on device dm-1, logical block 42500
Oct 16 06:55:46 pve-tr4 kernel: sd 0:0:0:0: [sda] tag#24 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_OK cmd_age=45s
Oct 16 06:55:46 pve-tr4 kernel: sd 0:0:0:0: [sda] tag#24 Sense Key : Not Ready [current]
Oct 16 06:55:46 pve-tr4 kernel: sd 0:0:0:0: [sda] tag#24 Add. Sense: Logical unit not ready, hard reset required
Oct 16 06:55:46 pve-tr4 kernel: sd 0:0:0:0: [sda] tag#24 CDB: Write(10) 2a 00 01 25 40 38 00 00 08 00
Oct 16 06:55:46 pve-tr4 kernel: EXT4-fs warning (device dm-1): ext4_end_bio:353: I/O error 10 writing to inode 5637299 starting block 42503)
Oct 16 06:55:46 pve-tr4 kernel: Buffer I/O error on device dm-1, logical block 42503
Oct 16 06:55:46 pve-tr4 kernel: EXT4-fs warning (device dm-1): ext4_end_bio:353: I/O error 10 writing to inode 5768643 starting block 4652177)
Oct 16 06:55:46 pve-tr4 kernel: Buffer I/O error on device dm-1, logical block 4652177
Oct 16 06:55:46 pve-tr4 kernel: ata1: EH complete
Oct 16 06:55:46 pve-tr4 kernel: Aborting journal on device dm-1-8.
Oct 16 06:55:46 pve-tr4 kernel: ata1.00: detaching (SCSI 0:0:0:0)
Oct 16 06:55:46 pve-tr4 kernel: EXT4-fs error (device dm-1) in ext4_reserve_inode_write:5941: Journal has aborted
Oct 16 06:55:46 pve-tr4 kernel: EXT4-fs (dm-1): ext4_do_writepages: jbd2_start: 11263 pages, ino 5636746; err -30
Oct 16 06:55:46 pve-tr4 kernel: EXT4-fs error (device dm-1): ext4_journal_check_start:84: comm pmxcfs: Detected aborted journal
Oct 16 06:55:46 pve-tr4 kernel: EXT4-fs error (device dm-1): ext4_journal_check_start:84: comm kworker/u131:1: Detected aborted journal



actualy i run with kernel 6.8.12-13 pined but want tu upgrade

have somebody a idea from problem

thank's

 

[fiona]

Hi,
what does smartctl report about the disk health? If all looks fine there, it might be a kernel regression. You could try the newer opt-in 6.17 kernel to see if it's fixed there.

 

[denut]

hello

thank's

fiona​

for you answer

kernel 6.17..2-1 tested
same problem after 10 minutes web interface not rushable



kernel 6.8.12-13-pve pinned again

waiting for a solution

 

[fiona]

What exact disk model do you have? Are there firmware updates available?

 

[jimmy1987]

Could it be you have a broken disk?

 

[momus]

It may be powersaving on sata line. Search this forum, there is topic about it, you probably must add option to grub to disable powersaving on sata line. Try add this: ahci.mobile_lpm_policy=0, if not help, try this: libata.noacpi=1, remember to run, update-grub after change

 

[denut]

It may be powersaving on sata line. Search this forum, there is topic about it, you probably must add option to grub to disable powersaving on sata line. Try add this: ahci.mobile_lpm_policy=0, if not help, try this: libata.noacpi=1, remember to run, update-grub after change

hello thank's verry much

look like working
powered on since 2h no more problem
with the "ahci.mobile_lpm_policy=0" option added to grub and kernel Linux 6.17.2-1-pve

have a nice day !!!