How do you interpret Seagate's SMART data? Signs of failure?

[speedboxx]

I've always had Western Digital drives, but recently I picked up a Seagate 500GB 7200.12 drive for cheap. I also have another 80GB Seagate EE25 in my car.

With Western Digital, the SMART seems to be intuitive and if the read error rate is non zero, then it could indicate you are having a problem. With Seagate however, the read error and hardware ECC recovered values are always high and always increasing. But from my understanding this is normal. Im guessing it is normal for a drive to have read errors all the time, but as long as it is corrected by ECC then you are fine (WD probably covers this info up?).

What SMART values do you guys usually pay attention to that may indicate a drive is failing? Just the current pending sector and reallocated sector values? Strangely, I had a WD drive that had a current pending sector of 50+, but after I wrote zeros to the drive and formatted, that value went back to 0...so I am unsure how to interpret this.

[andyb]

This is good info for all drives.

Reallocated_Sector_Count

Anything other than 0 is bad, but how bad depends on when the drive got them, and how long ago, if the last bad sector was reallocated 2 years ago you are probably OK now, as the problem has appeared, been fixed and has not come back. If you get a bad sector, you will likely get more, and the vauge statistic is that a drive that starts to get bad sectors will be trash within 6-months.

Reallocated_Event_Count

This can be a little confusing, you may get an "event", but the sector is never actually reallocated, this is usually due to a problem reading/writing a sector, which has then been flaged as bad, but when the drive got around to reading/writing to it at a later date it was OK, and therefore not actually reallocated. This means that you can get a "Reallocated_Event_Count" that is higher than the "Reallocated_Sector_Count", this can usually be ignored unless you have some of the above and some of the following.

Current_Pending_Sector

This is simple, sector(s) are awaiting a further look to see if there really is a problem (there usually is). I often see these when windows is broken, and the drive already has a few Reallocated Sectors. If so, the drive is usually on its way out, remember the 6-month rule.

Offline_Uncorrectable

The drive is very likely in a bad way, but the only way to tell is to write data to the whole drive and see what happens.

UDMA_CRC_Count

This can be a serious issue and is usually due to cabling faults, there is no damage to the drive usually swapping out the data cable fixes this problem. If you have swapped the data cable there are only a couple of other options, your motherboard or HDD are causing "soft faults" that could be an indication of worse things to come. If you have a laptop, there is rarely a cable or adaptor that can cause any kind of data transmission faults so it wcould be an issue with the drive or the laptop.

Most of the rest are not very serious, and the above are the only ones I ever look at (which I do with every single drive, whether the HDD is a suspect or not, the amount of times I have looked at a machine that obviously has viruses etc AND bad sectors, at which point the viruses are usually not a problem as the drive probably needs to be replaced.


Andy

[lobuni]

With Western Digital, the SMART seems to be intuitive and if the read error rate is non zero, then it could indicate you are having a problem. With Seagate however, the read error and hardware ECC recovered values are always high and always increasing. But from my understanding this is normal. Im guessing it is normal for a drive to have read errors all the time, but as long as it is corrected by ECC then you are fine (WD probably covers this info up?).

Yes these errors occur on other drives also. They just don't report them the same way. For my old 80GB Seagate HDD the results are:

Code: Select all

  Raw Read Error Rate  61  112478456  Good 
  Spin Up Time  99  0  Very good 
  Start/Stop Count  100  24  Very good 
  Reallocated Sector Count  100  0  Very good 
  Seek Error Rate  85  378843895  Very good 
  Power On Hours Count  87  11922  Watch 
  Warning: Power On Hours Count is below the average limits (88-100). 
  Spin Retry Count  100  0  Very good 
  Power Cycle Count  97  3238  Watch 
  Warning: Power Cycle Count is below the average limits (98-100). 
  Hardware ECC Recovered  61  112478456  Good 
  Current Pending Sector  100  0  Very good 
  Offline Uncorrectable Sector Count  100  0  Very good 
  Ultra DMA CRC Error Rate  200  0  Very good 
  Write Error Rate  100  0  Very good 
  TA Increase Count  100  0  Very good 

Notice that even after many years the Hardware ECC Recovered == Raw Read Error Rate.

[Wibla]

This is actually quite an interesting field, because the various manufacturers implement SMART in different ways.

I'll give some examples from my own seagate LP drive, one that is slowly failing (reallocated sectors increasing...), but still shows as OK in smart.

The interesting parts pertaining drive failure from this smart data sample is field 5, Reallocated_Sector_Ct, the raw value is at 58, but 0-255 value is still at 99 - with the "optimal" being 100, this means the firmware doesnt deem this to be very serious - yet, the treshold value is 36. At that point it will scream about drive failure, and rightly so. Just a few days ago reallocated sectors was 52, so its definently deteriorating.

Also you can see from the graph that several values fluctuate over time, unlike for example samsung drives, especially you can see Hardware_ECC_Recovered, Seek_Error_Rate and Raw_Read_Error_Rate going up and down depending on drive load.
Oddly enough raw_read_error_rate usually stays above 100, with "treshold" value being 006, apparently its operating "above optimal"(?)...

Code: Select all

=== START OF READ SMART DATA SECTION ===
SMART Attributes Data Structure revision number: 10
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate     0x000f   114   099   006    Pre-fail  Always       -       75441011         <--
  3 Spin_Up_Time            0x0003   100   100   000    Pre-fail  Always       -       0
  4 Start_Stop_Count        0x0032   100   100   020    Old_age   Always       -       12
  5 Reallocated_Sector_Ct   0x0033   099   099   036    Pre-fail  Always       -       58               <--
  7 Seek_Error_Rate         0x000f   068   060   030    Pre-fail  Always       -       6502443          <--
  9 Power_On_Hours          0x0032   098   098   000    Old_age   Always       -       2019
 10 Spin_Retry_Count        0x0013   100   100   097    Pre-fail  Always       -       0
 12 Power_Cycle_Count       0x0032   100   100   020    Old_age   Always       -       12
183 Unknown_Attribute       0x0032   100   100   000    Old_age   Always       -       0
184 Unknown_Attribute       0x0032   100   100   099    Old_age   Always       -       0
187 Reported_Uncorrect      0x0032   100   100   000    Old_age   Always       -       0
188 Unknown_Attribute       0x0032   100   100   000    Old_age   Always       -       0
189 High_Fly_Writes         0x003a   098   098   000    Old_age   Always       -       2
190 Airflow_Temperature_Cel 0x0022   079   074   045    Old_age   Always       -       21 (Lifetime Min/Max 17/21)
194 Temperature_Celsius     0x0022   021   040   000    Old_age   Always       -       21 (0 16 0 0)
195 Hardware_ECC_Recovered  0x001a   042   039   000    Old_age   Always       -       75441011         <--
197 Current_Pending_Sector  0x0012   100   100   000    Old_age   Always       -       0
198 Offline_Uncorrectable   0x0010   100   100   000    Old_age   Offline      -       0
199 UDMA_CRC_Error_Count    0x003e   200   200   000    Old_age   Always       -       0
240 Head_Flying_Hours       0x0000   100   253   000    Old_age   Offline      -       117763708291019
241 Unknown_Attribute       0x0000   100   253   000    Old_age   Offline      -       2415043908
242 Unknown_Attribute       0x0000   100   253   000    Old_age   Offline      -       741141060

[zsero]

Be very careful with that 7200.12, I have a 30% rate of them falling into bad sectors. On many forums, the 7200.12 series is now one of the NO-NO drives these days!

[speedboxx]

Crap...after the 7200.11 fiasco, I was hoping Seagate would have improved their 7200.12 line. One thing that's good about the 7200.12 compared to the similar WD Blue is that there are less platters on the 7200.12...which I would have thought increases reliability. However, the Seagate doesnt have the touchless head load/unload like most modern drives since it's still a contact start/stop.

[andyb]

During the 7200.11 fiasco we had to return 8x 1.5TB drives in one go, we simply got a refund and bought Samsung drives instead. we have been using nothing but Samsung SATA Desktop drives since that point, and have had very few problems.


Andy