Sign in with
Sign up | Sign in
Your question

New RAID1 already failed?

Last response: in Storage
Share
December 8, 2009 9:48:44 PM

Hey all, I had a HDD go down on me for the first time ever last week. The drive was a little over a year old, so it's death surprised me as I've used drives for 4-5 years in the past without fail. I have an external backup that runs daily, so I didn't lose much, but I have been re-installing everything since last Wednesday.

To avoid doing that all over again, this time I've installed a RAID1 setup using the Intel drivers for my ICH9R southbridge. Last weekend I started hearing the "cla-click" noise that was the death knell of the previous drive. Upon booting, it goes through the RAID check routine and says Status: Degraded and the second drive says "error occurred." Now within Windows, the system has 3-4 seconds of hang every few minutes, presumably from trying unsuccessfully to write to the redundant drive.

So it seems I've killed a second drive in a week, whereas I've never had a drive die on me before. I'm leaning towards a problem outside the drives themselves, like the SATA ports or the overall power just isn't enough. Has anyone experienced anything like this? If I do return the broken drive, can I copy the living one to the backup and keep a RAID setup, or am I starting all over again? Thanks for the help.

SETUP:

Case: Antec Sonata III
MoBo: Gigabyte EP45-UD3R
HDDs: 2x Seagate ST310005N1A1AS-RK
RAM: 2x 2GB Mushkin something or other (don't remember offhand)
DVD-RW: Pioneer 216D
OS: WinXP SP3
Vid: NVidia GeForce 6600

More about : raid1 failed

a b G Storage
December 9, 2009 4:43:43 AM

Well, it's just possible that your drive has given up the ghost. I assume it's still covered by warranty, since the standard for HD manufacturers is 3 years?
m
0
l
December 9, 2009 4:12:24 PM

That's a good thought; the old one is only about 18 months old.

But I'm confused as to how I've managed to kill two hard drives in two weeks... My guess would be either a motherboard problem or a PSU problem. Deductive reasoning, I know, as those are the only two connectors on it.

Would it be possible for me to replace the redundant drive with the same model and still keep a RAID setup without having to re-install everything I just did? Using Norton Ghost or something? It seems like the idea behind RAID would allow for replacing a dead drive and putting another in without skipping a beat.
m
0
l
Related resources
December 9, 2009 5:15:17 PM

Unless you're only using a single spindle (no RAID) don't use Seagate AS drives.

I've had a terrible problem with reliability of Seagate 7200.10 and 7200.11 drives lately. Specifically, they fail out of RAID all the time. Seagate doesn't say much of anything about this in the public literature, but after multiple RMAs, I contacted them directly and their RMA "Engineer" said that the AS drives were not designed for, and would not be reliable in "Enterprise RAID" which he defined as RAID 5 or RAID 10.
When I pointed out that the RAIDs it was failing out of were on desktops, which these days come with on-board RAID controllers almost universally, he continued to insist that what I was doing was "enterprise".
It's well documented that there are firmware issues, but the real problem is that seagate is selling a product that isn't suitable for a large, and growing, segment of the market: anyone using RAID, and not making it clear that the product won't work reliably in that configuration.
m
0
l
December 9, 2009 6:18:59 PM

Well, that totally blows. I needed a hard drive quick, as the office comp I built (TY forum help on that) went down last week, so I ran across town to Best Buy. I guess I can get a replacement online now. Any suggestions?

Also, could I disconnect the failed drive to get rid of the 3-second hang I'm getting? I know the redundant drive is already corrupt, so would this harm the system in any other way?

Lastly, I assume I'd need to make an image of the drive if I plan on putting different drives in the array. I've only used about 100GB of space on the 1TB drive. Am I going to be making a 1TB image, or will I only need space for the 100GB I've used? I've got a 500GB external that could hold the smaller image.

Again, thanks to everyone for their help.
m
0
l
December 9, 2009 10:04:52 PM

I've found that often the only way to get the system stable is to disconnect the failed drive. If you disconnect the drive that is already marked offline, things shouldn't get worse, but I can't guarantee they will get better.

I don't know what imaging software you plan on using, so saying how much space it will take is hard to gauge. Some will require like sized drives, others only the used data space. Funnily enough, Acronis disk image (an OEM version of which ships with Seagate retail drives) seems to do a good job of taking up only the used, versus the raw, capacity.
m
0
l
December 22, 2009 5:08:31 PM

So I got two new HDDs, and used the Acronis software that came with the Seagates to make a backup image. I then installed the new drive, which worked perfectly, but it wasn't a member of the RAID setup. To make it part, I would have to delete everything off of it.

I then went the other way, and installed the second drive in place of the failed drive. In the RAID setup, I made it a member, which said the setup would finish within windows. I couldn't find the Intel Matrix Storage Manager software, so I tried downloading and installing it. Upon installation, it told me there was a more recent version already installed. I've searched in my Program Files folder, but I can't seem to find it.

Would it be easier to put both new drives in, install Windows and create a RAID array with those, then install the backup image to the new drive? It just seems I'm stuck without the Intel software.
m
0
l
!