6

I am using a Seagate Backup Plus 4TB drive attached over USB to a Linux box. The drive model is ST4000DX000-1CL160.

When checking the S.M.A.R.T. attributes I get:

    $ sudo smartctl -a -d sat /dev/sdb
    smartctl 5.41 2011-06-09 r3365 [i686-linux-3.2.0-37-generic] (local build)
    Copyright (C) 2002-11 by Bruce Allen, http://smartmontools.sourceforge.net

    === START OF INFORMATION SECTION ===
    Device Model:     ST4000DX000-1CL160
    Serial Number:    Z1Z046LE
    LU WWN Device Id: 5 000c50 04ec317ca
    Firmware Version: CC44
    User Capacity:    4.000.787.030.016 bytes [4,00 TB]
    Sector Size:      512 bytes logical/physical
    Device is:        Not in smartctl database [for details use: -P showall]
    ATA Version is:   8
    ATA Standard is:  ATA-8-ACS revision 4
    Local Time is:    Sat Feb 16 13:55:29 2013 CET
    SMART support is: Available - device has SMART capability.
    SMART support is: Enabled
    ...
    Vendor Specific SMART Attributes with Thresholds:
    ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
    ...
    190 Airflow_Temperature_Cel 0x0022   040   039   045    Old_age   Always   FAILING_NOW 60 (0 209 61 41)
    ...
    194 Temperature_Celsius     0x0022   060   061   000    Old_age   Always       -       60 (0 20 0 0)
    ...

At first I thought that the failure of attribute 190 (Airflow Temperature Celcius) was caused by the drive not being in the smartmontools database and therefore being misinterpreted.

However, on second thought, the temperatures reported look sensible: 60 degrees celcius for the drive and 40 degrees celcius for the airflow. However, the threshold for attribute 190 seems incorrectly encoded. I guess the 45 degrees is the maximum allowed airflow temp. However smartctl will complain when the current value is below the threshold which it is now.

Can smartctl be instructed to interpret the threshold as a maximum rather than a minimum so it no longer displays an error?

mgd
  • 267
  • 1
  • 3
  • 9
  • 60 is the actual temperature, not 40. – Michael Hampton Feb 16 '13 at 19:54
  • I believe the threshold isn't in units of C. It is a normalized value. But, I too am getting failing smart stats based on Seagate's own smart normalized values and thresholds. –  Jun 15 '13 at 02:11

2 Answers2

11

Just an update for anyone Googling for this attribute, Airflow_Temperature_Cel values are read by SUBTRACTING the reported value from 100. So, for the report:

ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH
190 Airflow_Temperature_Cel 0x0022   040   039   045

The threshold is actually set at 55 deg C

Current value is 60 deg C

And the worse recorded value is 61 deg C

These values align with those reported in ID#194 Temperature_Celsius

TonyB
  • 134
  • 1
  • 2
  • Thanks for taking time to reply to this old question. The drive is dead long ago, but I learned something new. Could you add links to references to your answer, please? – mgd May 19 '17 at 14:41
  • While this interpretation of the smart values happens to work for this drive, it is generally incorrect - VALUE, WORST and THRESH are arbitrary values chosen by the manufacturer. Most drives with this attribute, however, report the actual temperature as raw value, so this can be used to get an idea of how these values interact. – Remember Monica Nov 23 '21 at 10:51
  • Many drives also keep a historical temperature log that can be read via smartctl -l scttemp – Remember Monica Nov 23 '21 at 10:51
  • And lastly, I have a large collection of various seagate and wd external disks (dozens), and the seagate disks can usually be brought to smart failure by a single long selftest, while the western digital, under identical conditions, usually stay at 55°. Seagate disks clearly have an overheating problem. – Remember Monica Nov 23 '21 at 10:53
5

You can make smart ignore the reading of the value by adding -I 190 -i 190 to the command line in your /etc/smartd.conf config file (See that page). Not sure it affects smartctl behavior though.

Generally, 190 can be ignored as indicator. But, 194 should not and your disk appears to be pretty hot, if you want to preserve it, you should ensure it gets colder. I invite you to read this research paper from Google regarding disk lifetime (with a paragraph for temperatures).

Just in case, you can also run Seagate tests on your disk (if possible), to ensure it's fully healthy SeaTools.

WinEunuuchs2Unix
  • 103
  • 4
Heis Spiter
  • 598
  • 7
  • 17
  • Thank you for the answer. I must admit that 60 degrees Celcius is very hot. However, the drive is an external drive (attached to a Linux box) standing in room temperature (20 degrees) with free air in all directions and it is only a little hot on the outside (around 30 degrees) and brand new. So it might be a sensor error or a misinterpretation of the SMART data by smartmontools (on Linux). I will test with SeaTools. – mgd Feb 17 '13 at 21:49
  • btw, as for my original question. Is there a way to tell smartctl how to interpret the temperature value or change the threshold? – mgd Feb 18 '13 at 07:12
  • If smartctl is not affected by the changes in config file, I don't think so. – Heis Spiter Feb 19 '13 at 07:37
  • Correct. The file `smartd.conf` only affects `smartd`. However, I wanted to know whether it is possible to change the thresholds and interpretation of temperature values. Are you aware of that? – mgd Feb 19 '13 at 15:44
  • Changing thresholds, no. Changing interpretation, this is doable through [`-v` argument](http://smartmontools.sourceforge.net/man8/smartctl.8.html) of `smartctl`. Even though I'm not sure it can help here. – Heis Spiter Feb 19 '13 at 16:40