NVM disappers

Samsung SSD 990 PRO 4TB. Machine with Proxmox. During long backup of VM Proxmox host hangs. After reboot no NVMe device detected.

• Reinsertion of NVMe in M.2 slot helps
  • related to temperature ?
• Multiple reboots doesn't help even if device is cool.
• If NVMe device disappears only complete power off helps.

...
nvme nvme0: I/O tag 642 (8282) opcode 0x2 (I/O Cmd) QID 10 timeout, aborting req_op:READ(0) size:131072
nvme nvme0: I/O tag 642 (e283) opcode 0x2 (I/O Cmd) QID 10 timeout, aborting req_op:READ(0) size:131072
nvme nvme0: Abort status: 0x0
...

Related:

• Problem with NVME timeout and aborting

Seems to help:

• nvme_core.default_ps_max_latency_us=0 pcie_port_pm=off pcie_aspm=off

  • Also hangs but after long time.
• disable ASPM in BIOS