TrueNAS HDD temps

[LordShadowen]

Just updated to 5.0.5. Also read https://github.com/rcourtman/Pulse/blob/main/docs/UNIFIED_AGENT.md#smart-disk-temperatures where it mentions S.M.A.R.T. Disk Temperatures,
Are the temperatures of HDDs (spinning rust, not SSDs) supposed to be supported now? Is there some manual action that needs to be taken?

If not, are there plans to support it? And ultimately, with proper Threshold support? My ultimate goal would be to get alerted if the HDDs are getting too hot...

[rcourtman]

S.M.A.R.T. disk temperatures (including HDDs) are supported, but require pulse-agent running on the TrueNAS host itself. Some clarifying questions:

1. Are you running TrueNAS SCALE (Linux) or CORE (FreeBSD)?
2. Is the agent installed on TrueNAS? Check with systemctl status pulse-agent on SCALE.
3. Is smartmontools installed? Check with which smartctl.

For TrueNAS SCALE, install the agent via Settings > Agents > Install New Agent, then run the generated command on TrueNAS.

TrueNAS CORE (FreeBSD) is not currently supported since there is no FreeBSD binary.

[LordShadowen]

1. TrueNAS Scale
2. Yes, version 5.0.5, confirmed with systemctl status pulse-agent. I do get everything else (CPU, Mem, etc...). Also ps aux | grep pulse confirms it's running as root
3. Yes - running sudo which smartctl returns /sbin/smartctl. Also, running e.g. sudo smartctl --all /dev/sde returns the complete info of the drive, including temperature.

In the UI, the Hosts section shows this:

Also tried reinstalling the agent, but nothing changed.

[pulse-triage[bot]]

Could you try enabling debug logging on the agent and check if S.M.A.R.T. data is being collected?

Step 1: Enable debug logging by adding to /etc/systemd/system/pulse-agent.service under [Service]:

Environment=LOG_LEVEL=debug

Then:

systemctl daemon-reload
systemctl restart pulse-agent

Step 2: Check logs for SMART-related messages:

journalctl -u pulse-agent -f 2>&1 | grep -i smart

You should see either:

• Collected S.M.A.R.T. disk data with a disk count (working)
• Failed to collect S.M.A.R.T. data with an error (problem with smartctl)
• No SMART messages at all (code path not executing)

Please share the output so we can diagnose further.

[LordShadowen]

I've upgraded to 5.0.7 and it started working, even before upgrading the agent (which was a bit strange).
After upgrading the agent, still works - I now can see temperatures (great!)

But two issues:

1. I can't "see" all the disk temperatures - the detailed temps of each disk can only be seen on the "hover" style box that shows up when hovering the main temperature value. This box is taller than the page and get's cropped. Can only see the bottom values.
   In general, this type of hover box is not great - a details sub-section similar to what containers get in the UI would be much better, so that we can see the details of the host

2. The temperatures of the disks can now be viewed, but it's not something that get's alerted on. My goal was was to be able to specify threshold values for the temps and get alerted if they go over that value.