I finally figured out what was crashing the servers here at dvorak.org. Bad firmware on the new Seagate 1.5T Barracuda drives. Everyone makes mistakes but in this case there is no firmware to download. You have to email them and then wait till they reply. And to email them you have to fill out a long form to create an account and agree to their terms, and when you do it creates an account – but the account doesn’t work. Seagate used to be a good company but I am not as impressed as I used to be. In the mean time the drive has been moved to a less critical server.
Also – updates need to run under either DOS or LINUX. Not running WINDOWS on this server.
I have over one year, bought 4 of these drives:
Seagate 146GB Ultra320 SCSI Hard Drive, 68pin, 15K RPM, 1″ Height, ST3146854LW, OEM, Recertified
Price Ea: $115.00
With only one bad one, that gives me this message:
Drive on 29320A at slot 0301:08:00 scsi ID2 has exceeded Failure Prediction Threshold. Make backup
I’ve been using WD exclusively for at least five years. My problems seem to be mostly MBs burning out or the power frying everything. I’ve seen very few drive problems, especially considering how mechanically intensive they are.
The last HD I found bad was a Maxtor in a friend’s computer. Geeze, it sounded like a snow plow on a gravel road! My WDs? No problem so far.
*
Marc, I’m glad you seem to have isolated the problem. I wish you the best.
17, Oops, its maxtor..
It is a colossal failure of QA. They basically skipped QA for the firmware. Or they loaded the wrong firmware for all the automated tests.
You would think they would fix it ASAP.
For anything that looks like a drive failure, and isn’t a total hardware disaster like Marc’s bad flash, I recommend Steve Gibson’s Spinrite. It’s saved me countless hours of grief. What can it hurt to try and recover your data with the best recovery software available?
http://grc.com/sr/spinrite.htm
My three 1.5TBs are slow. Screw this, I’m going back to my Microscience RLL drive.
I stopped buying Seagate about 2 years ago when I had support problems. I bought a brand new drive. When I opened the box, it was a refurb. Never again.
Marc I feel your pain and I referred the pre-sales Seagate tech support person I am working with to this post for comment.
I am getting ready to buy 2 1TB versions of this drive for a computer but will obviously wait a bit and do some more investigating first. Their support person said they “had a problem but it’s fixed.” As Ronald Reagan said, “Trust but verify.” I have to do something because I am running out of space.
I recently had a fairly new 500GB WD My Book external from Costco start the clicking sound of death and lost about 8 months of all my pictures including several that were used by CNN and all of the raw files for my newspaper for the same period. I know it was my fault for not having a backup but every time I look at that drive I want to cry.
For $1,500 bucks a clean room might be able to fix it…..
Total PITA but Seagate is not alone.
I had no problems getting the firmware for my shiny new Seagate monster drive, and it arrived as an iso to burn a CD image. It’s a boot disc, it doesn’t care what flavor of OS you’re running. A few minutes later I had a drive that’s been issue free for over a month now. It’s a no-brainer.
Seagate has some pretty good looking help online (by way of Engadget).
Uh, sorry, but Gizmodo sez thet the firmware fix for Seagate Barracuda hard drives just turns them into paperweights.
I suggest people try the firmware released today:
http://seagate.custkb.com/seagate/crm/selfservice/search.jsp?DocId=207951
#43 – B. Dog, check this site out. http://storagesecrets.org/2009/01/seagate-boot-death-analysis-overhyped-fud/
According to this, the likelihood of SD1A bricking a drive is HIGHLY unlikely. However, you should still back up your data and update your drive to SD1B.
I’m still disappointed in the way Seagate handled this, but I suspect this will all soon be forgotten.
These drives are still better than Fushitsu drives.