PDA

View Full Version : Help!!!!! How do I save data from stripe set with a dying drive?????



kleima
09-18-2003, 11:37 AM
Help!!!!! How do I save data from stripe set with a dying drive?????

The drive is one of my newest 73 GB SCSI320 Seagates. It is part of a three drive stripe.
I hear the drive click like it is shutting down, and then after awhile it spins up like it does when booting. I can't copy large files to or from the drive at this point without it shutting down. Then I get a error saying the data could not be copied and was lost. In the disk manager it shows the stripe set as "at risk" and it shows one of the three with a yellow exclamation mark saying errors.
Priority #1 is to try to save some of the data.
Then, I don't know how to tell which physical drive to replace, because it doesn't identify the SCSI ID in the disk manager (that I can see).

wvp
09-18-2003, 01:30 PM
One thought would be to disconnect the data cable from one drive (can you open the case & hear which drive is doing this?) and see if the disk manager still shows the "problem" drive. Another thought would be to check seagates web site or adaptects or even western digitals for a utility that can check each one of the drives.
AS for saving the data - most people are using Raid 0, which puts a little bit a each file on all the drives connected (as I understand it). This makes getting the data difficult, as entire files are not kept on each drive. One solution would be to have a disk repair company retreive the data (expensive), another option is a program called R-Studio (http://www.r-tt.com) $80 that can reconstruct some raid setups.
If your data is that important, I would stop trying to access the drives until you get the data retrieved.
Jeff
P.S. The R-Studio software can make an image of the bad disk first & recover from that before the drive totally goes.

SBowie
09-18-2003, 02:58 PM
Originally posted by kleima
In the disk manager it shows the stripe set as "at risk" and it shows one of the three with a yellow exclamation mark saying errors.[/i] It may be too late, but sometimes you can right-click on the drive that has the yellow bang and select "Reactivate", then copy the data off onto some stable volume asap!

kleima
09-18-2003, 08:15 PM
Reactivate is greyed out on both the stripe set with the problem, and the other stripe set.
I'll see if I can get that program. Hopefully they have a downloadable version.

bradl
09-22-2003, 09:13 PM
I lost one (of 4) of my stripe drives today for the first time. It just died. Holy crap that is not a good feeling. I checked disk manager and the bad drive had a Red X (I think) on it. I rebooted and as the SCSI manager scanned the drives a briefly saw an error message as it finished booting up. Drive was still dead. Shut down, removed the side and unplugged and replugged the power and SCSI cable to every drive and MB connection, then said a little prayer, booted up and it came back. Sure got me to thinging about a disaster recovery plan and at the very least a spare drive on the shelf matching the others...

Hope you work it out OK,

Scott Bates
09-22-2003, 09:20 PM
Something's going around. I lost one of four drives in my IDE set last Monday and one of four on my SCSI set on Friday. Fortunately I have enough room on the sets to back up to each other and they didn't go down at the same time. Warranty replacements are "in the mail", but work flow slowed down considerably. :(

Steve Thompson
09-23-2003, 02:12 AM
Hey guys,

Not to one up anybody (four-up) but I lost two Quantums about a month and a half ago and then again one more after getting them replaced and then shortly after that my Seagate Cheetah died.

I bought the Quantums when I went to my first Raid Setup and put the Cheetah in, that I already had, to make up the fourth. The Quantums were only two years old and under warranty and I thought that I might be out of luck on the Cheetah but it was still covered (5 year warranty) whoo hoo!

Both Maxtor and Seagate handled the replacements friendly and efficiently but here's something that might help those that are new to warranty replacements.

Both companies want their hardrives packed just so in a static free bag with nothing attached, blah blah blah and shipped just so with an attached RMA number or the warranty will be voided.

This seemed to me like an easy out for not covering the warranty but that's OK because I stumbled across the ADVANCED RMA.

If this is not new to you then I apologize because this was a big headache saver for me.

Tell the company that you would like an advanced RMA. Here's how it works. You give them your credit card number and they ship you out another hardrive the next day; all correctly packed which usually arrives withing 2-3 days. You then put your old hardrive in the same shipping containers (packed the way they like) and ship it back to them within the companies timeframe (Maxtor 4 weeks, Seagate 2 weeks) or you've just bought a new hardrive.

Hope this helps someone who loses a hardrive or two or four.

Steve

Scott Bates
09-23-2003, 08:50 AM
Advanced RMA is what I normally do Steve, but just discovered it's not an option with IBM (through Hitachi) unless it's so deeply hidden on their web site that I couldn't find it. Have to send in the dead drive and they'll send a replacement within 14 days, if in stock.

Steve Thompson
09-23-2003, 09:38 AM
Hi Scott,

Actually, Advanced RMA's do seem to be a hidden option and it was never openly offered by either of the two companies I previously mentioned.

I was talking to a friend about the pain of having to send back my drives and the packaging them properly or else part of the deal and he mentioned the advanced RMA.

When both companies tried to give me an RMA to send back my product, I asked if they had advanced RMA and they seemed kind of shocked that I knew about this option.

I remember having to send my first BOB back to NEWTEK to get fixed and how much I hated being without it. I delayed sending it in because most of the inputs & outputs worked. I would have loved to have had a replaced BOB before I sent mine in.

BTW: NEWTEK was very quick on replacing my BOB so I'm not trying to dis them.

Anyway, from now on I'll definitely ask about this option. I think it's a great way to do business.

Steve

kleima
09-25-2003, 11:07 PM
Well, I'll give you all an update! I tried the R-Studio software. No go. I should have guessed that, since the drive kept shutting off and then spinning back up a few minutes later. But, what can I say, I was hoping for a miracle.
Finally, after calling about twenty data recovery places and getting estimates ranging from $800-$20,000, and being advised to NOT apply power to the drive anymore; I backed-up and cleaned off as much space on other drives as I could, said a prayer and powered up!
I copied off, for five hours, every single file of the 168 GB! (I thought I had only 110GB on there, so had freed up enough for that!) I filled every single drive to capacity, even the drives on the other computer, across the network! The thing wouldn't even run for five minutes, before.
So, I guess the prayer was the most effective!
Then, I checked disk manager, the stripe was still at risk! This time when I right clicked on the drive with errors, "reactivate" was no longer greyed out. So, I reactivated, and the stripe turned "healthy!"
So, do I trust it now?

Scott Bates
09-25-2003, 11:42 PM
A real good question Kleima. After losing one drive on my IDE stripe last week I've been running on the other three while waiting for the replacement to arrive. A couple days ago, I began playing back a project I had just played a few minutes before without problems and out of the blue I get some sort of a system read error message and the LED for the stripe set is on solid. Although the system was acting like everything was locked up I managed to properly shut down both VT and the system. When I restarted and checked Device Management it showed all three drives in the set offline. Since I suspected one particular drive I shut down again and pulled the cables from that drive. Restarted and although only two of the three drives were there, they were both shown as online again. I just left it alone since I had all the project files backed up on my SCSI set so I changed the paths in the VT[3] Spreadsheet and completed the project. After shutting down I plugged drive #3 back in and restarted - my stripe set was back and all the data was still intact! Ran Western Digital's drive test software on all three drives and they each passed all the tests. It's been running fine since and then today got the replacement drive and reset the stripe set with four drives again. So far so good. Can I trust it any more than you can trust yours?

lwaddict
09-26-2003, 11:08 AM
Fujitsu FinalData 2.0 is the way to go...
in the event of system failure you can still retrieve files.

I use several programs here for data recovery, I support
a large number of end users with critical data...that never
seems to get put on the network where it belongs.
So I'm in charge of data recovery for now as well as other things.

Anyway...
FinalData can recover your files and only runs about 79 bucks,
a far cry from what data recovery people will charge...often up
to 500 bucks just for trying, much more for succeeding.

Hope this helps.