HDD diversity

[john_v]

Hi,

I was thinking if this would be a good idea to to get different hard drives for 4-bay NAS (QNAP-439Pro to be specific) so they won't die all at once.

I was looking at:
1) WD 1TB Green WD10EADS
and
2) Hitachi 1TB 7K1000.B HDT721010SLA360 - 0A38016

the big difference between those two that prompted my question is the cache size between those two: 32Mb and 16Mb respectively.

hence is the question for you guys: is it a good idea to mix-and-match HDD like that? Any tips and/or suggestions?

Thank you!

 

[thiggins]

I don't know if your theory that they won't all die at once because they come from different manufacturer will stand up. Some folks swear that all the drives in an array must be the same, right down to the manufacturer lot. Others say you have to mix lots.

The only thing that I would recommend is that you check the NAS approved drive list to check that the drives are on it.

I don't think you'll see any difference in performance due to cache size.

 

[john_v]

Thank you for your reply.

I have TWO clients who lost their RAID because of that (years ago though). One had RAID-1 with two HDDs and another had RAID-5 with eight HDDs.
they both had same model, same batch HDDs. One hard drive died, and in less than a week, another one - crashing array in both cases.
I might not see something like that in my life, but this experience does not comfort me.
The HITACHI drive I mentioned is NOT on QNAP's list, but their tech support told me it will work with the latest Firmware (whatever their word worth).

thanks again for your time and for such informative web site.

John.

 

[thiggins]

When you say "crashing the array" do you mean that the data was lost? A single drive failure in either a RAID 1 or RAID 5 array should not cause data loss.

 

[john_v]

<snip> One hard drive died, and in less than a week, another one - crashing array in both cases.
<snip>
John.

no, two drives died within approximately a week of each other. I think one case was just 2 days apart, and another case was about a week. In both cases there was not enough time to DELIVER disks. no one suspected a second failure would happen that fast - would have walked to BestBuy and got a disk - only if the client knew....

 

[thiggins]

Ah. The best protection, however, would have been a backup...

Time is of the essence when you have a failed drive in an array. But you already know that...now!

 

[john_v]

well, so true...

back to where it all started - HDD manufacturer diversity. Just ordered QNAP TS-439Pro with two WD's and two HITACHI's drives. Once I put it all together, I will update this post with the results... (remember - HITACHI I bought is not on their official web site, but their tech support said it will still work). we shall see.

 

[john_v]

well, so true...

back to where it all started - HDD manufacturer diversity. Just ordered QNAP TS-439Pro with two WD's and two HITACHI's drives. Once I put it all together, I will update this post with the results... (remember - HITACHI I bought is not on their official web site, but their tech support said it will still work). we shall see.

well, here are the results (plus some extra observations):
1) two WDC drives in bays 1 and 2:

Hard disk model_______Western Digital Caviar Green family
Model number_______WDC WD10EADS-00M2B0
Serial number_______W -DMWVA05317777
Disk capacity_______931.51 GB
Firmware version_______01.00A01
ATA version_______ATA-8
ATA standard_______Exact ATA specification draft version not indicated


2) two HITACHI drives in bays 3 and 4:

Hard disk model Hitachi_______HDT721010SLA360 ST6O
Model number Hitachi_______HDT721010SLA360
Serial number_______TS6F40RM12JTPJ
Disk capacity_______931.51 GB
Firmware version_______ST6OA31B
ATA version_______ATA-8
ATA standard_______Not recognized. Minor revision code: 0x29


Everything seems to work together just fine. My only concern was that "minor revision code" under ATA standard. I've asked their support via MSN chat and I even opened ticket through their web site - no response. I expected a bit better handling from a company who makes NAS'es for business... but it is a separate story.
I've noticed that HDD's in bays 1 and 2 (WDC's) are constantly а few degrees cooler than HDD's in bays 3 and 4 (HITACHI):

System temperature 28°C/82°F
HDD 1 temperature 32°C/89°F
HDD 2 temperature 32°C/89°F
HDD 3 temperature 33°C/91°F
HDD 4 temperature 34°C/93°F

the HDD in bay 4 (the right most - by the wall) is always the warmest one.

you can feel the CPU power - for sure. the management GUI is fast.

I've also played with RAID expansion RAID-1 -> RAID-5, it worked flawlessly. it took about 4 hours to complete, but I was able to access the files on NAS. Only once, in the very beginning - I've lost the connection to the shared drive, but it came back upon refresh - almost immediately - not sure what it was.

everything else seems to be functioning as expected.... If anyone have any questions about it - please let me know.