I’ve been playing around with NVMe ssds, basically the laptop-style stuff with a M.2/NGFF connector.
Mine are Samsung 950pro and learn more I’ve also poked at them using nvme-cli.

There’s one thing that sticks out, on each shutdown the unsafe shutdown counter increased:

localhost:/# nvme smart-log /dev/nvme0 Smart Log for NVME device:/dev/nvme0 namespace-id:ffffffff critical_warning : 0 temperature : 50 C available_spare : 100% available_spare_threshold : 10% percentage_used : 0% data_units_read : 7831129 data_units_written : 8263496 host_read_commands : 34443111 host_write_commands : 108073938 controller_busy_time : 310 power_cycles : 23 power_on_hours : 608 unsafe_shutdowns : 17 media_errors : 0 num_err_log_entries : 1 Critical Composite Temperature Time : 0 Temperature Sensor 1 : 0 C Temperature Sensor 2 : 0 C Temperature Sensor 3 : 0 C Temperature Sensor 4 : 0 C Temperature Sensor 5 : 0 C Temperature Sensor 6 : 0 C Temperature Sensor 7 : 0 C Temperature Sensor 8 : 0 C

I tried to find out something more and there’s a command (protocol-wise) the SSD should be sent in a final step before shutdown.

https://communities.intel.com/community/tech/solidstate/content?itemView=thumb&filterID=contentstatus%5Bpublished%5D~category%5Bcompatibility%5D

I’m assuming this must be a kernel issue with the NVMe driver, but I’m not sure. It’s certainly not good though.

(from redmine: issue id 5082, created on 2016-02-07)

• Changesets:
  • Revision eef5ee88 by Carlo Landmeter on 2017-05-22T15:35:27Z:

community/nvme-cli: upgrade to 1.3 and move to community

refs #5082