My RAID5 Arracy died again!

[Paddy]

I have four 250GB Seagate drives conencted to an adaptec 2400A (I think) RAID 5 controller. I have had problems with faulty drives inthe past, so I have the adaptec monitor on startup which constantly reports on the status of the array. Yesterday morning everything was fine.

I got a call from Andrea at work to say that one of the drives had failed. At that point I should have made a backup of my documents. Instead I walked her through the process of rebuilding the array. When I got home I saw that the third disk had failed, but it was rebuilding. As it takes ages, I left it to it and went inside. I tried to stream some video from the array, which again was probably a bad idea, and after two mins the video crashed. I went back to the file server and it reported that the array was dead as two drives had failed! Bugger.

I have the Seagate drives with the five year RTB and they are on continuously. I find it hard to believe that both would fail on the same day. In any case I dont want to touch the drives too much. The RAID controller wont mount the array as it has a dead status, but I can see that all the drives are spinning and being detected. Is there any way to force the controlelr to mount the array so that I can get my data from it.

If not, is there any recovery software that will see the four drives so I can recover something. I should have learned by now I need to get access to about 50MB of docs.

Now the reason i am panicking is that this is my backup. My main PC is being rebuilt at the moment.

NOTE TO SELF. USE DVDR!

Any suggestions???

[Paddy]

I am working nights tonight, so during my day off I have brought the fileserver and main pc into the living room.

I've had a partial success. In moving the PC the second drive that had failed has fixed itself. Perhaps it was a power failure. In any case, the drive on bus2 has failed.

I should be able to use and rebuild the array as i have only one dead drive, but the Adaptec utility reports that the array is dead due to multiple drive failures. On boot, the controller shows the array. I just need to somehow 'reset' the flag so that it thinks that the array is degraded.

[Dr Mordrid]

I know this is herecy, but I avoid Seagate drives. Too many bad experiences.

[Paddy]

I think it is temperature related. I am cooling the drives now and the array is more stable. I can access the data albeit with difficulty. It looks likek two drives need to be replaced. It's a shame that Seagate don't advance ship

[Drizzt]

2 HP server
Each one had 3 Maxtor 250GB in Raid 5
One of the two had another Maxtor 250GB USB disk for backup.
Everything is protected by an UPS chain (4 top of the top UPS)


The same day, at a distance of a few minutes one from the other, 2 disks of each array AND the backup disk broke down.


Brought to a data recover service, discovered that:
1) All 5 disks had the internal "things" broke down in half.
2) In the last days they had to recover data from a wide number of 250GB disks: about 26 disks in 12-15 days...all broken in the same way









Side note: anyone has the same feeling I have, that disks are becoming more and more frail instead of improving?
Instead of having disks with less (the hell...what it is the name? Dish?) inside, spinning faster, with high data density, smaller and hotter...wouldn't be better to have bigger things with more (again, dishes? Plates?)?
I'd really love to have in servers 5,25'' disks, with 4-8 "dishes" (so to read more data without the need for 10k/15k rpm), maybe with 2 set of heads front-to-front parallel reading...

[Enak]

I think it is temperature related. I am cooling the drives now and the array is more stable. I can access the data albeit with difficulty. It looks likek two drives need to be replaced. It's a shame that Seagate don't advance ship

Seagate do advance ship. Well, they did the last time I looked.

[Taz]

again, dishes? Plates?

Platters

[Drizzt]

Platters

Like the singers?

[Paddy]

I don't think Seagate advance ship to outside the US.

I have rebuilt 75% of the array by chilling it Now I need to replace two of the disks, but obviously i can't return them without replaceing them first.

[degrub]

eh Paddy ! You need a couple spares anyway ! Goferit

[Paddy]

the drives are cheap enough. I'll buy a spare one and then start the RMA process. Since keeping the drives chilled and rebuilding the array I am detecting no faults.

I'm going to put them in a HTPC and improve the cooling.

[Fat Tone]

If you are building a HTPC, don't try and use a fancy low-profile case. I did (Silverstone LC11) and you have to jump through a lot of hoops. Stick to a desktop size with plent of room for quiet cooling and no need for risers.

[degrub]

i second that. Larger fans = more airflow = slower rpm = less noise. THe antec p150 is a slick looking quiet case if you can stand glossy white. i'm sure there are others.

what works for drives works for beer too !