nvme: Ubuntu uses nvme temp sensor on dashboard

[daberkow]

I figured it was worth mentioning, it seems the Ubuntu 26.04 default dashboard when you login shows temp, and having no where else to find it it gets it from the nvme driver.

Welcome to Ubuntu 26.04 LTS (GNU/Linux 7.0.0-28-generic x86_64)

 * Documentation:  https://docs.ubuntu.com
 * Management:     https://landscape.canonical.com
 * Support:        https://ubuntu.com/pro

 System information as of Fri Sep 11 02:27:31 UTC 2026

  System load:  0.08              Processes:               120
  Usage of /:   2.0% of 95.85GB   Users logged in:         0
  Memory usage: 3%                IPv4 address for enp0s8: 172.30.0.14
  Swap usage:   0%                IPv6 address for enp0s8: fd7a:a62:9c1d::7
  Temperature:  -273.1 C
root@ubuntu-pxe-server:~# cat /sys/devices/pci0000\:00/0000\:00\:10.0/nvme/nvme0/hwmon0/temp1_input
-273150

[iximeow]

I was curious what this looks like elsewhere and picking AWS as no particularly exciting comparison:

> ssh -i ~/nvme0/aws_3.pem ubuntu@35.88.62.128
Welcome to Ubuntu 26.04 LTS (GNU/Linux 7.0.0-1006-aws x86_64)

 * Documentation:  https://docs.ubuntu.com
 * Management:     https://landscape.canonical.com
 * Support:        https://ubuntu.com/pro

 System information as of Fri Sep 11 19:16:24 UTC 2026

  System load:  0.01              Temperature:           -273.1 C
  Usage of /:   30.7% of 6.61GB   Processes:             117
  Memory usage: 25%               Users logged in:       0
  Swap usage:   0%                IPv4 address for ens5: 172.31.37.186


Expanded Security Maintenance for Applications is not enabled.

0 updates can be applied immediately.

Enable ESM Apps to receive additional future security updates.
See https://ubuntu.com/esm or run: sudo pro status


The list of available updates is more than a week old.
To check for new updates run: sudo apt update
ubuntu@ip-172-31-37-186:~$ cat /sys/devices/pci0000\:00/0000\:00\:04.0/nvme/nvme0/hwmon0/temp1_input
-273150

so at least in that sense things aren't weirder here. I'm surprised that our banners have such different orders though!

one other detail here as I was looking, what does the SMART log say, and from AWS it is all zeroes like I expect you would see from us:

$ sudo nvme smart-log /dev/nvme0
Smart Log for NVME device:nvme0 namespace-id:ffffffff
critical_warning                        : 0
temperature                             : -273 °C (0 K, -459 °F)
available_spare                         : 0%
available_spare_threshold               : 0%
percentage_used                         : 0%
endurance group critical warning summary: 0
Data Units Read                         : 0 (0.00 B)
Data Units Written                      : 0 (0.00 B)
host_read_commands                      : 0
host_write_commands                     : 0
controller_busy_time                    : 0
power_cycles                            : 0
power_on_hours                          : 0
unsafe_shutdowns                        : 0
media_errors                            : 0
num_err_log_entries                     : 0
Warning Temperature Time                : 0
Critical Composite Temperature Time     : 0
Thermal Management T1 Trans Count       : 0
Thermal Management T2 Trans Count       : 0
Thermal Management T1 Total Time        : 0
Thermal Management T2 Total Time        : 0

that... makes sense to me. there aren't really useful values to report for anything here in the typical cases. for us: is Data Units Written useful when the underlying storage media is behind Crucible anyway? so all zeroes is kind of what I was hoping for. I think this is where the driver gets -273.15 from.

I happened to find on a 22.04 VM on Propolis that, uh, the tool does not report all zeroes:

$ sudo nvme smart-log /dev/nvme0
Smart Log for NVME device:nvme0 namespace-id:ffffffff
critical_warning			: 0
temperature				: -273 C (0 Kelvin)
available_spare				: 0%
available_spare_threshold		: 0%
percentage_used				: 0%
endurance group critical warning summary: 0
data_units_read				: 2595936641960814343328586490970112
data_units_written			: 2595936641960814343469323979325440
host_read_commands			: 2595936641961423086023756394528768
host_write_commands			: 2595936641960814343469323979325440
controller_busy_time			: 140726007342687
power_cycles				: 0
power_on_hours				: 0
unsafe_shutdowns			: 2583625040005507927782165772763136
media_errors				: 210453397504
num_err_log_entries			: 140058596018912
Warning Temperature Time		: 4294967295
Critical Composite Temperature Time	: 4294967295
Temperature Sensor 5           : -261 C (12 Kelvin)
Thermal Management T1 Trans Count	: 49
Thermal Management T2 Trans Count	: 60
Thermal Management T1 Total Time	: 4294967295
Thermal Management T2 Total Time	: 0

this is with nvme-1.16. on 24.04, with a much newer nvme-cli, it is all zeroes. it would be good to understand what happened with the log page here, since data_units_read and data_units_written for example sure look like pointers when you read them as hex. something here is super wrong with older nvme, and I'm going to track that down, but that's distinct from the -273.15c disk :)