smartd: What Warnings Do I Heed, Which Do I Ignore?

Question

I've enabled smartd on my Gentoo box, but now I've got a torrent of warnings in my logs and in Logwatch. My question is basically if these warnings should be dealt with, or squelched - and if there's a better way to monitor it than just logwatch?

Logwatch output:

 /dev/sda [SAT] :
   Usage: Temperature_Celsius (194) changed to
     116, 115, 114, 113,

 /dev/sdb [SAT] :
   Prefailure: Raw_Read_Error_Rate (1) changed to
     116, 117, 118,
   Usage: Airflow_Temperature_Cel (190) changed to
     74, 73, 72, 73, 72, 73, 72, 71, 70, 71,
   Usage: Hardware_ECC_Recovered (195) changed to
     62, 61,
   Usage: Temperature_Celsius (194) changed to
     26, 27, 28, 27, 28, 27, 28, 29, 30, 29,

 /dev/sdc [SAT] :
   Prefailure: Raw_Read_Error_Rate (1) changed to
     118, 119,
   Usage: Airflow_Temperature_Cel (190) changed to
     69, 68, 67, 68, 67, 66, 65, 66,
   Usage: Hardware_ECC_Recovered (195) changed to
     52, 53, 54,
   Usage: Temperature_Celsius (194) changed to
     31, 32, 33, 32, 33, 34, 35, 34,

 /dev/sdd [SAT] :
   Prefailure: Raw_Read_Error_Rate (1) changed to
     106, 107, 108, 109, 110, 111, 112, 114, 115,
   Usage: Airflow_Temperature_Cel (190) changed to
     70, 69, 68, 69, 68, 69, 68, 67, 66, 65, 66,
   Usage: Hardware_ECC_Recovered (195) changed to
     64, 65, 66, 65, 64, 65, 66, 65, 64, 62, 61, 60,
     59,
   Usage: Temperature_Celsius (194) changed to
     30, 31, 32, 31, 32, 31, 32, 33, 34, 35, 34,

/var/log/messages excerpt:

Dec  9 07:29:17 megatron smartd[5554]: Device: /dev/sdc [SAT], SMART Prefailure Attribute: 1 Raw_Read_Error_Rate changed from 106 to 108

Answer 1

In short, these seem to be mostly INFO level messages. I recommend that you take a look at this thread at linuxforums: http://www.linuxquestions.org/questions/linux-hardware-18/smart-prefailure-errors-on-new-hard-drive-645418/

and read a bit more about self-tests: http://sourceforge.net/apps/trac/smartmontools/wiki/FAQ#Selftests

Update Oh, and this http://www.linuxjournal.com/magazine/monitoring-hard-disks-smart?page=0,0 is a must read!