From: Bruce A. <ba...@gr...> - 2006-04-19 03:13:07
|
Jason, > i spoke with wd again and it is apparently some half implemented pcb > airflow sensor. I have been experimenting with 640 (no, not a typo) Western Digital disks on our new cluster (Nemo). They are: Model Family: Western Digital Caviar SE (Serial ATA) family Device Model: WDC WD800JD-08LSA0 I have determined that on these disks, Attribute 190 is a monitor of the MAXIMUM temperature ever detected by the disk. The raw value is the max temp in Celsuis. The failure threshold is set so that if the disk reaches 55 Celsuis (max operating temp) then the normalized value drops to the threshold threshold and the disk is marked as a 'usage failure'. Here is an example: ID# ATTRIBUTE_NAME VALUE WORST THRESH TYPE UPDA WHEN_FAILED RAW 190 Unknown_Attribute 064 044 045 Old_age Always In_the_past 36 194 Temperature_Celsius 107 088 000 Old_age Always - 36 In what follows I refer to Attribute 190. The raw value indicates that the CURRENT disk temperature is 36 Celsuis. The normalized 'VALUE' decreases by 1 for each 1 Celsius increase in temperature. The sum of 'VALUE' + 'RAW' is always 100. Thus the 'WORST' value of 44 indicates that at some time in the past the disk temperature was 56 Celsius. In smartmontools 5.37 I have changed the name 'Unknown_Attribute' to 'Temperature_Celsius'. The only difference with Attribute 194 is that Attribute 194 has its failure threshold set to zero so it can never fail. Cheers, Bruce |