Drive fails to join RAID group

[harpseal1]

Hey guys i need some input,

Ive got a 4TB SATA2 external raid device with 4 1TB Western digital drives. I am using the Silicon Image 3132 controller for this RAID (It came packaged with this MicroNet external raid)

Im trying to create a RAID 5 with these 4 drives, but drive number 4 always produces an error and the group creation is aborted. I pulled the drive from the array and used it as an internal drive in my workstation. The drive partitioned and formatted just fine. I ran checkdisk, no bad sectors or errors. I then loaded the drive with about 300GB of data and ran checkdisk again, the disk had some orphaned files and checkdisk was able to fix that...

So the drive appears to work fine other than the orphaned file problem. It just will not work with this RAID for some reason.

I switch the drive to bay number 3 in the micronet array and the error still appeared at that specific drive, so I don't think anything internally in the RAID controller is broken or a driver/software problem..

My question is, should i just RMA this "bad" drive to Western digital or what since it doesn't appear to be "trashed".

Error message from log file while drive is in bay 3:

RAID group/path/target 0:0:2, Sense Data 70:00:0B:00:0 (ScsiError)
Create aborted, Group 0
Member 2 (Path ID = 0 Target ID = 2 dropped from group 0 "RAID5"

Error in bay 4:

RAID group/path/target 0:0:3, Sense Data 70:00:0B:00:0 (ScsiError)
Create aborted, Group 0
Member 3 (Path ID = 0 Target ID = 3 dropped from group 0 "RAID5"

thanks, andy

 

[Onus]

Sounds like a bad drive to me; it orphaned files during your testing.

 

[retro77]

What OS are you running? If its 32-bit then you can not have a RAID array bigger than 2GB. I ran into the same issue with my 4 750GB drives. Instead of getting my full 3TB I can only have 2TB.

If you go 64-bit OS then you wont have this issue.

 

[harpseal1]

Oh i forgot to specify, im using windows xp 64 for these tests. But originally i had it on our 2003 server, which is 32 bit... So looks like ill just keep this controller in my workstation since i manage the stuff around here.

Thanks for the input guys, ill get that drive RMAed.

 

[kvalente]

Yeah, I'm having the exact same issue. Any solution?

 

[harpseal1]

Yeah Actually I did have a solution without having to send the thing back. Heres the deal:

I tested this thing for a week. I pulled out each individual drive and tested it, not one single bad sector or anything abnomal at all. I then flipped out drive 4, the original one getting kicked out during the raid creation with drive 3.

Guess what, drive 3, now in slot 4, was getting kicked out of the raid creation. Now i was getting a bit annoyed.

So i pulled the SATA card out and put it in an xp 64 workstation and setup the raid... Worked flawlessly.. It did take about 34 hours or something ridiculous for the raid to actually finish formatting and what not. I guess there is no quick format for a raid 5.

I did move the sata controller and the external drive back to the server, where they now work fine. I did notice that now the SATA raid manager says the raid 5 is in a degraded state, but the lights on the raid unit say everything is fine.

So then it dawned on me. There is another RAID controller in that server with a similar chipset. Basically, they were fighting and the raid configuration program was getting confused which controller was what and which drives went to it.

So, just keep that controller and its external raid in another machine that doesn't already have a RAID setup in it, although its perfectly fine as long as that machine has the motherboard raid controller not controlling a raid as well.