IT magazine reports 'Seagate ST31500341AS drives behave bad in big RAIDs'

[hawihoney]

For those interested the complete german article can be found here. The part about the RAIDs starts with "Anwender, die die 1,5-TByte-Platte ...":

 

http://www.heise.de/newsticker/Ausfaelle-bei-Seagate-Festplatten-durch-Firmware-Probleme--/meldung/121822

 

Seems that these cheap drives are only valid in desktop RAID systems (1-4 drives). For bigger RAIDs we should take the high-end drives especially build for this. Vibration in big RAIDs lead to the cheap drives being switched off by the RAID controller. This explains my 'high_fly_writes' on my 15* 1.5TB array. I do have the lastest firmware.

 

I thought this might be of interest.

 

Harald

 

[Joe L.]

For those interested the complete german article can be found here. The part about the RAIDs starts with "Anwender, die die 1,5-TByte-Platte ...":

 

http://www.heise.de/newsticker/Ausfaelle-bei-Seagate-Festplatten-durch-Firmware-Probleme--/meldung/121822

 

Seems that these cheap drives are only valid in desktop RAID systems (1-4 drives). For bigger RAIDs we should take the high-end drives especially build for this. Vibration in big RAIDs lead to the cheap drives being switched off by the RAID controller. This explains my 'high_fly_writes' on my 15* 1.5TB array. I do have the lastest firmware.

 

I thought this might be of interest.

 

Harald

 

On most RAID arrays, with data stripped across all the disks, there is a LOT of disk activity on all the disks with any "write" or "read" of data on the array.

With unRAID, this is not true.  Only the drive being read is active, or when writing, the parity drive and the drive being written.

 

When doing a parity check, there might be a bit more vibration, but since it is a linear read of each cylinder in turn on each disk I'd guess vibration is less than random seeks of a large number of drives.

 

Interesting article though...  It shows how tight the tolerances must be

• 
  
• 
  

in the higher density drives.