PERC H710 seems damaging my drives

Drives being swapped into a raid will have to be initialized and containers created.  You cant just willy nilly toss drives into a raid and it works.  All of these operations are done from the Raid Firmware not an os.

As unlikely as it sounds, it looks like you got bad drives. The PERC won't "damage" them, even if there was something wrong with it.

Did you also check the Advanced and Initialize boxes when you created the arrays (you initialize arrays, not individual disks)?

When you put them into "another PC", how did you determine that "they result bad drives"? If you took them straight out of the box and they are bad, whether on a server or PC, sounds like  you received defective drive(s).

I initialized (fast) the array, not individual disks.

When I put disks into another PC (Win764bit Pro), I did the following:

1) Checked if Win was able to detect disks. It detected both.

2) Try to create both a GPT or MBR partition table, create a partition and format it as NTFS: for one of the 2 drives I was able to create and format the partition, for the other one I wasn't even able to create the partition table.

3) Launched WD WinDLG ( support.wdc.com/.../download.asp;sid=3&lang=en ) and found the first drive as FAILED, the second one not even accessible by WD software.

4) Launched Both Quick and Extended tests on the only accessible drive, and it resulted with "FAIL: too many bad sectors detected".

Then I concluded the drives were failed.

I'm not one to say a thing is "impossible", but that is just about as close as it comes. The controller is not damaging your drives. The power doesn't come from the controller anyway - it comes from the backplane (or cables) from the system's power distribution nodes. Test a new drive in your PC first next time, so you know it's good before using, then put it in the server. Dell does not guarantee those drives will work in their server, but according to the specs, WD has tested it on an H710 to work.

Thanks.

Then you think it's really difficult that PERC H710 generated wrong volgages/currents able to kill my drives? I'm quite frightened to insert a new set of drives and test if they work...

Any electric-related known issue with that controller?

Agree with Flash, in 25 years never had a LSIlogic based adapter destroy a disk.

I just have ask.... were these NEW retail disks, where did you get them?

Why fear inserting another drive, almost all retailers warranty the drives

Thank you. Well, my fear is because if the problem is due to a controller/backplane/power supply electrical issue on the server that "destroys" my disks, I'm fear the disk retailer or manufacturer won't give warranty.
These drives have 5-years warranty.

Any way to troubleshoot electrical issues on the server without "killing" other drives?
I'm astonished if a so said "high-quality" server has no electrical protection system to avoid burning devices...
Thank you in advance.

Any way to troubleshoot electrical issues on the server without "killing" other drives? I'm astonished if a so said "high-quality" server has no electrical protection system to avoid burning devices...

There are many protections and monitors for such situations. You would get an error like E1000 FAILSAFE if there were such an issue. It just is extremely, EXTREMELY unlikely (impossible, really) for the server to cause the kind of damage you describe. Where did you buy them? Are you sure they are legit? You could always hire a PC electrician if you really think it is the server.

1. No, the OS could not do this.

2. See below and my previous posts.

I only asked about the disks because I have seen cheap Chinese knock-offs floating around - relabeled, etc.

It sounds also really unlikely that if the drive were unsupported by the controller (even if WD indicates it is), it would have been damaged in such a way. Do you agree?

Yes, I agree: It is really unlikely (impossible) IN ANY CONCEIVABLE SCENARIO (supported or unsupported) that it could damage the drives in such a way without getting some kind of warning on the server hardware. As I said, there are server errors that would be tripped (and the system shutdown and prevented from booting/POST) in such an event.

Thanks,
This is my real opinion:
1) The OS can't damage disks as it needs to "bypass" 2 firmwares (i.e. the PERC firmware and the WD firmware) to be able to execute activities as low level as needed to damage a disk. I think for an enterprise class disk like my WD4000FYYZ also a bugged PERC firmware shouldn't be able to damage the disk.

2) Electrically damage a disk is still possible if the data interface or power interface on the backplane supplies wrong currents/voltages to the disks interfaces. One who wants to sizzle a disk can supply an higher power than that supported by the interface or probably also an out of tolerance variable  power.

Disks are new and bought from amazon, I hope they're legit.

It sounds also really unlikely that if the drive were unsupported by the controller (even if WD indicates it is), it would have been damaged in such a way. Do you agree?

Thank you in advance.

Really easy to test voltage on a server if you have a volt ohm meter.

pcsupport.about.com/.../power-supply-test-multimeter.htm

Power anomalies due to power utilities or lightning can mess with disks, but they generally occur between 1am and 6am and are rare ( due to switching on the grid). Building machinery can cause them rarely, but not enough to effect disks (generally).

" I'm fear the disk retailer or manufacturer won't give warranty. These drives have 5-years warranty.

Do you rally think retailer/manufacturers test returned disk as to the cause of failure? Possible, but an electronics Engineer with a high degree of training in the area of chip design would be needed, plus many hours, not likely. Been returning disks under warranty for >28 years, never had a denial of claim.