SOLVED Bug? with Smart Temperature Reporting

[NugentS]

Hi All,

Not sure how to categorise this. Is it a bug, a feature, or that's the way it is?

Anyhow. The issue is with smart temperature and alerts.
Under services/s.m.a.r.t I can set some temperature thresholds that seem to apply to all the storage devices. This works.
The thing is that all storage devices are not created equal.
For instance - I have a DOM that boots freenas. This has different temperature thresholds than an HDD - yet the temp warning seems to apply to all equally.
Under storage/disks I have edited the device, enabled S.M.A.R.T. (I think irrelevant) and set the temperature thresholds for this device which are different from the HDD thresholds.

I still get warnings about the device using the general thresholds. I don't know if the disk specific settings are meant to override the general settings - but maybe they should.

 

[Samuel Tai]

If you examine /usr/local/etc/smartd.conf, you should see the device-specific temperature thresholds. However, SMART will alert on the most conservative thresholds, as it should.

 

[NugentS]

The numbers in smartd.conf seem correct and represent the general numbers and overrides I have entered.
Yet the alert is being generated on the wrong numbers for the specific device.

Sample below:
/dev/da0 -a -d removable -n never -W 10,34,38 -m root -M exec /usr/local/libexec/smart_alert.py\
-s S/../../(2|3|4|5|6)/(03)\

/dev/da0 -a -d removable -n never -W 10,34,38 -m root -M exec /usr/local/libexec/smart_alert.py\
-s L/../../(1)/(00)\

/dev/ada3 -a -d removable -n never -W 10,40,50 -m root -M exec /usr/local/libexec/smart_alert.py\
-s S/../../(2|3|4|5|6)/(03)\

/dev/ada3 -a -d removable -n never -W 10,40,50 -m root -M exec /usr/local/libexec/smart_alert.py\
-s L/../../(1)/(00)\

 

[Samuel Tai]

... the alert is being generated on the wrong numbers for the specific device.

Could you be more specific? From your smartd.conf, the lowest thresholds are 34 and 38 C. As these are the most conservative, smartd will alert on these regardless.

 

[NugentS]

ada3 is a silicon device - SATADOM
da0 is an HDD

ada3 is reporting warnings based on the 10,34,38 numbers rather than the 10,40,50 numbers

Sean

 

[Samuel Tai]

How does da0 report?

 

[NugentS]

Normally. The results from the HDD's are as expected
ada3 is basically an SSD it just runs hotter for whatever reason but is alerting on the HDD values. Its not a major issue - I am just a bit confused as to why. I don't really care what temp it runs at (unless it melts) as I am far more interested in the HDD's

 

[Samuel Tai]

Like I said, SMART will alarm on the most conservative values for all disks, under the belief that the temperature inside the case is mostly uniform.

 

[NugentS]

Fair enough - I am more interested in the HDD's anyway. I'll mark this one down as "that's the way it is"