Kudos to Hitachi

[Cave Man]

I have a 3TB Hitachi drive that failed a year and a half into its 3 year warranty. Got an RMA online, sent it in and two days after it arrived at Hitachi I have already received shipment notice that the new drive is going out on Monday. I only had crappy service from Western Digital, so hopefully they won't screw up the Hitachi brand now that they own them.

 

[brentsg]

I bet you receive a refurbished Western Digital replacement though.

 

[Cave Man]

That would be a dagger in the heart.

 

[brentsg]

Please post when you receive it. I've read of this problem at other forums. People had Hitachi drives fail out of a RAID array so they needed the same drive to put back in. However they received a WD replacement and had to go source a new Hitachi, which is getting both difficult and expensive at this point.

 

[murphychris]

Please post when you receive it. I've read of this problem at other forums. People had Hitachi drives fail out of a RAID array so they needed the same drive to put back in. However they received a WD replacement and had to go source a new Hitachi, which is getting both difficult and expensive at this point.

Well... while it's obviously preferable to get an identical model drive, if these are consumer drives, the manufacturers rather clearly say the only RAID supported is 1 and 0, anything else all bets are off. And for either RAID 1 or 0 you can mix and match drives.

For something like a NAS running mdadm software RAID 5 or 6, or ZFS RAIDZ then it still likely doesn't matter, because those RAID implementations are really tolerant of consumer disks taking a while to error correct, before dropping them out of an array.

The problem comes with either motherboard (integrated) RAID which can be much fussier. With higher end RAID controllers (sometimes called hardware RAID) people should be using SAS drives anyway, and for sure you're going to get an exact replacement as that's the whole point of getting an enterprise SAS drive - along with their vastly faster error correction so they don't get dropped out of an array.