help with SMART results

[dima]

results from smart check do not show any failures as i can see
- drives are at 20,000hr life (not much)
- smart overall/extended checks passed
- no crc errors

in the block of "SMART Attributes Data Structure revision number: 1" it mentions somewhat of a "pre-fail" as i see but not sure how to read these and i was not able to find info on-line
- ID 170 (Available_Reservd_Space) has flag as "pre-fail", but raw value is at zero (why, and is it something to pay attention to)
- ID 175 (Power_Loss_Cap_Test) has flag as "pre-fail", raw value is at 622 (1 1190); not sure what it may mean
- ID 184 (End-to-End_Error) has flag as "pre-fail", but raw value is at zero (why, and is it something to pay attention to)
- ID 232 (Available_Reservd_Space) has flag as "pre-fail", but raw value is at zero (why, and is it something to pay attention to)



(this is from Linux system)

if anyone can help with this will be greatly appreciated!!!

 

[dlavigne]

Were you able to determine if the drives had errors?

 

[AlexGG]

in SMART, there is VALUE and RAW_VALUE

VALUE is to be interpreted as "higher means better". It starts from 100, meaning "perfect" (or sometimes 200 or 250) and goes down as the drive condition deteriorates. If VALUE gets below THRESH, the drive is considered faulty. If THRESH is not specified (zero), the attribute is not considered critical.

RAW_VALUE is an internal source data, firmware-dependent, with generally no specified meaning (although common conventions do exist). Firmware will take RAW_VALUE, and produce VALUE based on it, and VALUE is then supposed to be used for diagnostic purposes.

In spinning drives, RAW_VALUE is useful when watching for bad sectors and some other things, and for some attributes will specify the actual number of bad sectors or whatever events. The list most often includes (from memory and in no particular order) Current Pending Sector Count, Reallocated Sector Count, Offline Uncorrectable Count, Reallocation Event Count. Obviously there are some others depending on a drive.

RAW_VALUE is not guaranteed to be useful, and some drives will even provide zeros at all times, while computing VALUE from some internal data.

Specifically to your case,

End to End error, RAW_VALUE is the number of errors (0), corresponding to VALUE of 100 (perfect, no errors), related to internal CRC error detection, inside the SSD itself.

Power Loss Cap Test, RAW VALUE is interpreted as 110 microseconds to discharge, last test performed 2 minutes ago, according to https://www.intel.com/content/dam/w.../ssd-power-loss-imminent-technology-brief.pdf and the VALUE is 100 meaning "perfect".

Available Reserved Space, apparently RAW_VALUE is always zero and VALUE shows percentage of reserved space available.

 

[Jessep]

in SMART, there is VALUE and RAW_VALUE

VALUE is to be interpreted as "higher means better". It starts from 100, meaning "perfect" (or sometimes 200 or 250) and goes down as the drive condition deteriorates. If VALUE gets below THRESH, the drive is considered faulty. If THRESH is not specified (zero), the attribute is not considered critical.

RAW_VALUE is an internal source data, firmware-dependent, with generally no specified meaning (although common conventions do exist). Firmware will take RAW_VALUE, and produce VALUE based on it, and VALUE is then supposed to be used for diagnostic purposes.

In spinning drives, RAW_VALUE is useful when watching for bad sectors and some other things, and for some attributes will specify the actual number of bad sectors or whatever events. The list most often includes (from memory and in no particular order) Current Pending Sector Count, Reallocated Sector Count, Offline Uncorrectable Count, Reallocation Event Count. Obviously there are some others depending on a drive.

RAW_VALUE is not guaranteed to be useful, and some drives will even provide zeros at all times, while computing VALUE from some internal data.

Specifically to your case,

End to End error, RAW_VALUE is the number of errors (0), corresponding to VALUE of 100 (perfect, no errors), related to internal CRC error detection, inside the SSD itself.

Power Loss Cap Test, RAW VALUE is interpreted as 110 microseconds to discharge, last test performed 2 minutes ago, according to https://www.intel.com/content/dam/w.../ssd-power-loss-imminent-technology-brief.pdf and the VALUE is 100 meaning "perfect".

Available Reserved Space, apparently RAW_VALUE is always zero and VALUE shows percentage of reserved space available.

Great post.