When did my drives get Hot?

[Johnny Fartpants]

Hi All,

I logged into the FreeNAS GUI today and noticed an alert saying a couple of my drives had got hot (above 41C) but annoyingly I wasn't emailed an alert although I did have this configured. Any ideas how I can track back and find out exactly when this happened?

I'm running FreeNAS-9.3-STABLE-201604150515.

Thanks

 

[anodos]

Hi All,

I logged into the FreeNAS GUI today and noticed an alert saying a couple of my drives had got hot (above 41C) but annoyingly I wasn't emailed an alert although I did have this configured. Any ideas how I can track back and find out exactly when this happened?

I'm running FreeNAS-9.3-STABLE-201604150515.

Thanks

try smartctl -x

You could also scroll back through logs to see if there was a time of unusually high system activity.

 

[Johnny Fartpants]

Thanks andos.

smartctl -x didn't help but managed to pick it up in the logs. I also noticed the failed to send mail error explaining why I didn't get a notification but don't understand how to fix (see below).

[common.system:210] Failed to send mail: [Errno 9] Bad file descriptor

I get all other notifications like updates, security run outputs etc just not S.M.A.R.T. Settings configured in the Services Tab.

Any ideas appreciated.

 

[DrKK]

smartctl -x didn't help

You did smartctl -x /dev/adaN for the appropriate value of N, and it didn't help? Are you sure? How about if you pastebin.com that output for us, and let us have a look?

Also, under the SERVICES menu, click on SMART properties. One of the final properties is "email address to report SMART problems". YOu must set that. It is separate from the other warnings (I am sure there is a reason for this, but, I don't know what it is).

 

[Johnny Fartpants]

You did smartctl -x /dev/adaN for the appropriate value of N, and it didn't help? Are you sure? How about if you pastebin.com that output for us, and let us have a look?

Also, under the SERVICES menu, click on SMART properties. One of the final properties is "email address to report SMART problems". YOu must set that. It is separate from the other warnings (I am sure there is a reason for this, but, I don't know what it is).

Yeah I have this setup as you describe and it used to work but no more even when I try to add different email addresses.

Thanks for your help.

 

[DrKK]

As I requested, would you please pastebin the result of smartctl -x {device}?

 

[Johnny Fartpants]

Sorry.

http://pastebin.com/qX5sWpWH

 

[Sakuru]

Huh, interesting. That's an unusual output. It's missing the "Vendor Specific SMART Attributes with Thresholds:" section. How about -a instead of -x?

 

[Johnny Fartpants]

Sure.

http://pastebin.com/dWwEQmPC

Its a SAS drive so maybe that's why?

 

[Sakuru]

Its a SAS drive so maybe that's why?

Yeah, that's probably it. That's unfortunate because other drives have a temperature history section that I believe DrKK and anodos were looking for.

 

[titan_rw]

Looks like -l scttemp is ATA only.

-x on a sas drive is equivalent to:

--health -i -A -l error -l selftest -l background -l sasphy

(no scttemp)

Even with sata drives that do support scttemp, the default logging interval is pretty short. 5 mins or something. So the entire temp history is overrun in a few hours.

Unless you change the temp logging interval that is, but very few people do.

 

[cyberjock]

Even with sata drives that do support scttemp, the default logging interval is pretty short. 5 mins or something. So the entire temp history is overrun in a few hours.

That's why I keep telling people that -x is stupid to ask for. It gives you a bunch of data that is useless unless you're quick on the draw. For 99.9% of us, but the time you see the overtemp condition, post on the forums asking for help, get a response, then do the command the info you need is expired.

 

[Ericloewe]

That's why I keep telling people that -x is stupid to ask for. It gives you a bunch of data that is useless unless you're quick on the draw. For 99.9% of us, but the time you see the overtemp condition, post on the forums asking for help, get a response, then do the command the info you need is expired.

Besides the temperature, it occasionally has some tidbits that help understand what's going on with the drive. That said, it's pretty rare to see -x bring something to the table that -a does not - since -a is more compact (and long enough as it is...), I tend to ask for that.

 

[anodos]

It'd be nice if the webgui had a hard drive temperature graph. Actually, what I really want is a VR GUI that I can manipulate like in Minority Report. Maybe I should make a feature request?

 

[Ericloewe]

It'd be nice if the webgui had a hard drive temperature graph. Actually, what I really want is a VR GUI that I can manipulate like in Minority Report. Maybe I should make a feature request?

Yes, I'm totally acquiring a Kinect 2 for finger tacking and a VR headset from dubious-quality HTC or evil-facebook oculus so I can use an unergonomic interface while making a fool out of myself to anyone who would happen to pass by.

Does my lack of enthusiasm for that idea show?

 

[anodos]

Yes, I'm totally acquiring a Kinect 2 for finger tacking and a VR headset from dubious-quality HTC or evil-facebook oculus so I can use an unergonomic interface while making a fool out of myself to anyone who would happen to pass by.

Does my lack of enthusiasm for that idea show?

Mostly your lack of imagination. :p

 

[cyberjock]

It'd be nice if the webgui had a hard drive temperature graph. Actually, what I really want is a VR GUI that I can manipulate like in Minority Report. Maybe I should make a feature request?

There has been multiple feature requests for this. The short reason is "no, we don't care enough to write the code, but if someone offered a pull request on github that worked, we'd consider it".

 

[anodos]

There has been multiple feature requests for this. The short reason is "no, we don't care enough to write the code, but if someone offered a pull request on github that worked, we'd consider it".

And here I was trying to make a joke thinking 'surely no one would actually want this'. :(

 

[danb35]

And here I was trying to make a joke thinking 'surely no one would actually want this'. :(

I assume @cyberjock was responding to the "temperature graph" part (which has been frequently requested), not the VR environment part.

 

[anodos]

I assume @cyberjock was responding to the "temperature graph" part (which has been frequently requested), not the VR environment part.

That makes more sense. I wrote my response immediately after finishing a ~15 hour drive and so my logicker was in a frazzled state.