SMART Service HD Temperature -> Email trouble shooting?

[BillyBob2]

My SMART service is not working properly. Or Maybe just the emailing component. Any help would be greatly appreciated.
in both "services" and on the disks them selves I setup smart monitoring. (I activated the SMART service on the disks as well)
in the SMART services I set the threshold to 10 degrees centigrade and to check every 1 min, and i have yet to receive an email. While the disks are running at ~30 degrees c

running 3 x 2TB Hard Disks plugged directly into my MB's sata ports. (MSI P67A-C43 (B3)) (4GB kingston DataTraveler G3 USB Flash Disk for the OS)
if I type smartd -d from the command line. All is looking good and displays good feedback regarding the status for the hard disks and there temps.

but not so fortunate from FreeNAS itself. I am not getting any emailed reports (System email SMTP is configured properly and test mail received).

What additional info do you need from me???

I am not seeing any specific error messages on the console, where would I see those if there are any?

If I stop the SMART Service, and then start it up again, then I get this:
" Feb 23 14:25:31 freenas notifier: smartd not running? (check /var/run/smartd.pid).
Feb 23 14:25:31 freenas notifier: Starting smartd. "

 

[joeschmuck]

Did you read the manual? Grab the PDF copy and read page 201. This tells you how to set it up. I suspect you forgot one key item. Keep in mind that you need to setup your email in FreeNAS also, so ensure you can send a test email and get it.

 

[cyberjock]

Not to mention that the threshold is the required change in temp between checks to receive an email. Since you check it once a minute(hint: this is way to often.. I'm not sure why you changed it but the default is a very good thing to leave alone) you'll NEVER EVER hit a chance of 10C in 1 minute.

If your hard drive starts heating up that fast it will probably liquify in your computer before you'd even notice you got an email. :P

 

[BillyBob2]

Joeschmuck & cyberjock, thank you both for the reply. Its very much appreciated.

Joeschmuck, LOL, I liked the good humor

The only reason I set it to 10 degrees and to check every 1 min, was to quickly test if the email is working....

so with all of that in mind. How do I know if the email is working or not????

As stated in my initial post, "(System email SMTP is configured properly and test mail received)"


but no emails from the SMART have been received. how do I know if the SMART monitoring alerting is working or not?

and thank you again

 

[cyberjock]

If the SMART service is running and you receive the test emails then it is working. With SMART, if you get any emails, thats a bad thing. So the fact that you haven't gotten any is a good thing.

 

[joeschmuck]

So you get test emails without trouble.

Do this if not already done...

1) Settings --> Advanced --> Show console messages in footer (checked)
2) Storage --> View Disks --> Verify "Enable S.M.A.R.T." has a value of true in each drive that you want to monitor.
3) Services (Left Pane) --> S.M.A.R.T. --> set to the following:
Check Interval: 3
Power Mode: Idle
Difference: 3
Informal: 20
Critical: 25
Email to report: (your email address)
4) Services (button) --> S.M.A.R.T. set to ON. If already set to on then turn it off for 30 seconds and then back on.

Assuming your drive temps are above 20C then in the footer window you should see alarm messages. If you do not see those messages then either SMART is not supported or I missed something in the setup. If the messages are there yet you do not see the email, check your email address.

If you are not getting the error messages and you believe SMART is running, open an SSH window and send the entire results of the following command (place in code wrapper t retain format), use the proper drive info for the example of /dev/ada0: smartctl -A /dev/ada0

The above steps for for me on FreeNAS 8.3.0-Release. You didn't mention what FreeNAS software you were running.

 

[BillyBob2]

Its working! Thank you, thank you, thank you! ( am on FreeNAS 8.3.0 p1)

I'm just a little confused as to why your settings worked vs mine.

The only thing I changes was as follows:
3) Services (Left Pane) --> S.M.A.R.T. --> set to the following:
Check Interval: 3 (was at 1)
Power Mode: Idle (was at Sleep)
Difference: 3 (was at 1)
Informal: 20 (was at 10)
Critical: 25 (was at 20)

Maybe the computer just needed to sit over night or an additional reboot.
I shut it down last night, and then turn it back on to try your settings.

 

[BillyBob2]

One more minor related issue.

I only received the email message the 1st of the two email address that I had listed for the notification email address.
I double checked that the email address's are entered correctly... divided by " ; " without spaces

 

[joeschmuck]

Confirmed Failure

Good Find!
Did you notice any error messages popping up? In your email was the body of the text empty? Mine was when I tested it.

Below is one group of error messages I had.

Code:

Feb 24 09:46:16 freenas smartd[6276]: Device: /dev/ada4, Temperature 31 Celsius reached critical limit of 27 Celsius (Min/Max ??/31) 
Feb 24 09:46:16 freenas notifier: joe@aol.com: not found 
Feb 24 09:46:18 freenas smartd[6276]: Warning via mail to joeschmuckatelli@aol.com;joe@aol.com produced unexpected output (34 bytes) to STDOUT/STDERR:  
Feb 24 09:46:18 freenas smartd[6276]: Null message body; hope that's ok 
Feb 24 09:46:18 freenas smartd[6276]: Warning via mail to joeschmuckatelli@aol.com;joe@aol.com: failed (32-bit/8-bit exit status: 32512/127)

For now only use one email address, I have reported the error and maybe it will be fixed in 8.3.1-Release. Also, those are not my real addresses in the error message.

-Joe

 

[joeschmuck]

Update,
This problem should be fixed in 8.3.1-Release. Looks like this was flagged 2 months ago and fixed. The problem is a comma should be used to separate the email addresses for smartd to accept it.

-Joe

 

[BillyBob2]

whoops! you are correct.
message body is empty.


One more related question (not an issue)

What if I want to get a daily digest of all of my drive temps in one email? In addition, maybe filtered SMART stats, but only if there was any issues found with a particular disk. (not enough to set off an alarm, but maybe minor issues? So as to spot the trend for future bigger potential issues.)

What do you think? Feature request?


Thank you again for you help and feedback

 

[ProtoSD]

whoops! you are correct.
message body is empty.


One more related question (not an issue)

What if I want to get a daily digest of all of my drive temps in one email? In addition, maybe filtered SMART stats, but only if there was any issues found with a particular disk. (not enough to set off an alarm, but maybe minor issues? So as to spot the trend for future bigger potential issues.)

What do you think? Feature request?


Thank you again for you help and feedback

Sounds like a great idea that I'm sure many others would like also. It would require a database to store the SMART results for comparison. Probably in a 9.x release it could be added.

Open a ticket at support.freenas.org and suggest it.

 

[ThomasDK81]

Thanks for the tips in this thread.
Does your HDDs really run so cool?
Normal temps for my 8x3TB WD Red is 35-36C. Of Course the don't ever idle.

 

[BillyBob2]

no, like yours in the 30 to 40 range.

I just set it that way to see if the SMART emails were being sent properly.

turns out there is some bugs in the system.

if you want to be notified to multiple emails, you should seperate with ":" not ";"

in addition to the emails being title properly, but there contents were blank.
this was a new bug now discovered...

so for now just send to single email, util the bugs are fixed in the next release.

 

[joeschmuck]

whoops! you are correct.
message body is empty.


One more related question (not an issue)

What if I want to get a daily digest of all of my drive temps in one email? In addition, maybe filtered SMART stats, but only if there was any issues found with a particular disk. (not enough to set off an alarm, but maybe minor issues? So as to spot the trend for future bigger potential issues.)

What do you think? Feature request?


Thank you again for you help and feedback

You could write a script to do that and it actually sounds easy, just break it down into segments.

Example (off the cuff):
1) Determine how often you want to sample the drives, lets say every 30 minutes starting at 00:01 and ending at 23:31.
2) Setup up a script to run every 30 minutes that would do the following:
a) Run smartctl and filter out the temperature for each hard drive.
b) Store the temperatures into a separate file defined by date and drive.
c) Set a flag if any one temperature is above the threshold.
3) Run a script at 23:45 to check for a set flag and if so create and format an email and send it.

You could also send the email immediately when a flag is set, what I'd prefer to do.
(If your good you will also store this data into a spreadsheet format and send it off as an attachment so you can see it in a chart or format the email so it displays a chart.)

The problems come in when you start using add-on SATA cards, smartctl doesn't always play well with those so it may be a custom script for those items. I'm trying to work something now in that arena but since I don't know python, django, or dojo it will be an uphill battle for me. My goal is to place it in the GUI and have all the options to display the temps over time and toss out an email as well. And to top it off, make it a plugin (that should be fun to try and make work:(). Just don't expect anything from me for a few months, I don't get much spare time to play around on the computer and I have a limited idea of how I'm going to accomplish this.

Hope this helps.

 

[BillyBob2]

I LIKE!

would be awesome if it was built in...