NVMe Ubuntu Controller reset


I am running ubuntu 20.04, with the OS running on a Mushkin 500GB SATA SSD. I installed a Samsung 970 NVME 500GB into the Motherboard M.2 2280 slot (ASRock H40M Pro4). The system found the device, which I was able to partition, format with xfs, and mount to /opt.

Within minutes of using the drive, it will unmount, and the device disappears from /dev. When I check the logs, I find the following:

Sep 23 19:07:12 vhost01 kernel: [12919.190730] nvme nvme0: controller is down; will reset: CSTS=0xffffffff, PCI_STATUS=0x10
Sep 23 19:07:12 vhost01 kernel: [12919.242746] nvme 0000:01:00.0: enabling device (0000 -> 0002)
Sep 23 19:07:12 vhost01 kernel: [12919.243096] nvme nvme0: Removing after probe failure status: -19
Sep 23 19:07:12 vhost01 kernel: [12919.266771] blk_update_request: I/O error, dev nvme0n1, sector 2048 op 0x0:(READ) flags 0x1000 phys_seg 1 prio class 0
Sep 23 19:07:12 vhost01 kernel: [12919.266847] XFS (nvme0n1p1): SB validate failed with error -5.
Sep 23 19:07:12 vhost01 kernel: [12919.292504] nvme nvme0: failed to set APST feature (-19)

I don't know if this is a motherboard problem, NVME device problem, or an issue with the kernel. The controller reset occurs randomly, but happens soon after I start accessing the device. It's happened when simply mounting the disk. However, the controller reset does not occur while the device is idle.

  • The motherboard is running the latest BIOS
  • The NVME stick is at the latest firmware
  • Tried using BtrFS over xfs, but the problem persists
  • smartctl -a /dev/nvme0 shows no problems or errors

Any idea what is causing this issue?


asked on Server Fault Sep 24, 2020 by rdege

0 Answers

Nobody has answered this question yet.

User contributions licensed under CC BY-SA 3.0