I hope this the right place to post this on.

It had to happen. I have never experience hard drive failure. I have just purchased a 1 TB Western Digital FW/USB-2 D to back up my files. My machine is three years old and has worked great. It has two Hitachi SATA 160GB intertnals, one Maxtor FW 120 external. They all work great. Two years ago I purchased another internal Western Digital WD3200/320 GB for more storage. This is the one that crashed. It spins up but the MOBO just does not see it. I do not remember how much is on the drive except that it is precious to me, all my veg files and some more storage.

A friend (guru) says find another exact drive and rob the logic board a try running it with the borrowed logic board. Sounds scary to me. The external Maxtor (F) is 99% full and the one of the Hitachi drives(D) is 99% full. The other Hitachi(C) is the System drive. This is why I bought the 1 TB for back up. I had just copied the files off the Maxtor and the one full Hitachi OK. The next day I was going to copy the files off the WD 320(E) but it didn't appear. Damn the luck.

I don't speak computers very well and thought I would ask you guys for suggestions before I did anything.

Your thoughts and any help is appreciated.

thanks,

Thomas Edward Bufkin

First of all, if there is a problem with the disk, and the material is precious, for a price you may be able to hire a service to recover the files. They can do a lot with broken drives, but it comes at a cost.

Second, I would go into BIOS setup to see if the drive is picked up in the primary set up menu. If that shows nothing, you might try shut it down, and try to attach the 320 to a cable to one of the drives you know is working, then restart, go into the Bios Set up again, and see if the drive is recognized. If it is, you may be okay.

Third, in older Windows versions, I recall you had to give the system permission to add new disks, and provide a range of names. I don't think that exists anymore, but I am wondering if the new terabyte storage disk, bumped the 320 off the list, in some way.

If you really aren't literate in this area, and this disk means something to you, I would get the computer to a repair facility. Under no circumstance should you reformat the disk if asked. You overwrite whatever is on there, if you accept that option.

put the drive in a plastic bag and put the bag into the freezer for a night.
the reconnect the drive immedaitely (while still frozen) to the PC (an external usb box is great for this).
80% of the data i saved were saved by this process.
The drive could work for a while but not for long, so be prepared to save the most important first.
swaping electronic is another trick, but finding the same drive can be tricky (unless you are always buying drives by pair like me).

Phew Thomas very sorry this ain't going to help, but you've just justified my expensive purchase of a WD 2TB My Book external drive to archive important footage. It's set up on Raid 1 to mirror the stuff.

Best of luck, please let's know how you get on thanks.

If you do set up a RAID array, be sure to check the files once in a while. I have heard about people not realizing their data was becoming corrupted over a period of months since the RAID up the errors.

One danger with RAID is that it isn't 100% foolproof. Things like data corruption, lighting / electrical surges, theft, fire (and to some degree user error) will take out both sets of data at once.

If you do set up a RAID array, be sure to check the files once in a while. I have heard about people not realizing their data was becoming corrupted over a period of months since the RAID up the errors.

I never heard this argument before and it makes me wonder why large corporations, or even small ones rely on for instance HP EVA configurations for their storage. If you can give me some solid examples you may convince me, but my initial reaction is that this is complete BS. I find it hard to believe that with storage capacities of up to 120 TB per 8100 system, you need to check files regularly.

Having heard without proof is IMO hearsay, always to be doubted and taken with a lot of grains of salt.

This should be handled automatically by any setup worth the rediculous money RAIDs cost. For example, my 3ware card verifies data automatically, comparing to the parity information, at least once a day. The card manufacturer recommends if it finds a data mismatch more than 3 times a month, to replace the drive in question.

I've had one drive go out since I installed the RAID. It gave over 100 errors for data/parity mismatch in a week (all of which pops up tons of alerts to the desktop with the 3ware software), and was still hanging in there until I could replace it! By the time I replaced it it was showing up as yellow in the drive software, yellow being to 'watch' the drive, not even red which was a drive 'warning'. Kinda similar to the hurricane/tornado watch/warning system. Possibility vs imminent. :)

There should be settings to make sure that it is checking the info when you're not using the computer, or even when you are. It'll do it slowly so as not to have an impact on performance.

Carl

Carl,

I do not agree that Raids are expensive. A 16 disk SATA2 raid controller with 2 GB cache memory is only around $ 1.000 for the best brand available. How does that compare to the price of Adobe Production Premium of around $ 1.900? Adobe software needs to be upgraded every two years for around $ 800 each time, while your raid controller goes on and on.

The investment in the raid controller gives you up to 10 times the performance of your disks in comparison to single disks.

The cost of raids is in the disks if you go SAN, not in the controller. For SATA it is not expensive.

Glenn,

If you look at the HP XP24000 array with up to 1152 disks, 851 TB storage and up to 247 PB capacity (yes, that is Peta bytes) and 256 GB cache, the stories you heard make even less sense. These are very popular arrays and widely used. No way to check the files regularly on those kinds of storages.

The HP StorageWorks XP24000/XP20000 Disk Arrays are large enterprise-class storage systems designed for organizations that simply cannot afford any downtime or tolerate any data loss. The XP mitigates the risk of business downtime by providing a bulletproof platform with complete hardware redundancies, hot-swappable components, and non-disruptive online upgrades. Data replication and tightly integrated clustering solutions, along with disaster recovery support, enable a multi-site disaster tolerant design to achieve complete business continuity. And with enhanced data protection and security features, you can decrease exposure to data loss.

Hi Harm,

1- Take my comments with a little grain of salt since backups are not my thing.

2- I think what people are most interested in are backup strategies for their particular situation... i.e. considering their budget and their needs. The HP solution is probably more robust due to battery backup, clean power, security systems in the data center, etc. But that probably does not apply to the original poster.

It may be slightly better to go with separate drives stored in separate locations, or a tape backup solution. (Tape backup probably makes a lot of sense if you are shooting on a data camera like the HVX, Red, etc.)

I know somebody who has had 2 drives fail at once in their XSERV RAID, losing their data on it. (Yikes.) So I wouldn't count on RAID 5 being a particularly bulletproof solution.

3- Buy anyways... if you have some good suggestions for backup for what most folks on this board are doing then I would be interested. Right now my backup strategy is not the greatest (I have some data that is not backed up... bad me).

Glenn,

The way I do my backups is the following: My documents and data are made available off line, so that is one copy. These are automatically synchronized on log on and log off. The online documents and data are available on the server in raid5, 2-nd copy (or should I say the original). All documents and data are automatically copied daily thru a VPN connection to another server, again in raid5 in a different location. In addition for particularly sensitive data, I make copies to an external disk.

All original video tapes are stored separately, all end results after editing are written back to tape and stored and of all produced DVD's one DVD copy is saved as well.

One drawback of RAID is that it does alter the mean time to failure for the array.

Thomas, I've been there done that and I have used Norton Ghost several times to pull the old files off of a "dead " drive...
If it will spin up then you have a chance... try to access the drive with Norton Ghost and when you see the files pull them off of that drive onto another known good drive...

you wont be able to image the drive such that it boots or such but you can get your files off of the drive...

One drawback of RAID is that it does alter the mean time to failure for the array.

That is correct but is no problem, or at least very easily remedied with the proper equipment. For arguments sake, let's assume you have an 8 disk array in raid5. Of course the mean time to failure is shorter than with a single disk. Now a disk breaks down and you get the degraded array message. The only thing you need to do is check the leds to identify which disk is faulty, hot swap the disk for a new one and the controller will rebuild the array in the background. Problem solved. No downtime, no data loss.

If your data is really critical, you have options for other raid configurations, like raid6 or raid10, so you can have multiple disks breaks down at the same time without losing data and without downtime. But the main drawback of not having a raid, missing the redundancy, is that a disk failure is catastrophic, causing downtime and loss of data.

I never heard this argument before and it makes me wonder why large corporations, or even small ones rely on for instance HP EVA configurations for their storage. If you can give me some solid examples you may convince me, but my initial reaction is that this is complete BS. I find it hard to believe that with storage capacities of up to 120 TB per 8100 system, you need to check files regularly.

Having heard without proof is IMO hearsay, always to be doubted and taken with a lot of grains of salt.

Here is a case (http://photo.net/bboard/q-and-a-fetch-msg?msg_id=00NV9g):

...these pictures were stored on a primary hard drive and backed up to a mirrored RAID array. Unfortunately, when primary hard drive crashed I found that the RAID had been silently corrupting files for a year - but of course, without having a digest or checksum of each file and comparing them regularly, this could not be detected. I also had some CDs that went bad.

Your case does not prove your point. The post you quoted also said he had CD's go bad on him. This points more to where a large majority of errors occur, between the chair and the keyboard.

It does not mention hardware or software raid, nor the raid configuration, apart from mirroring, so it could be raid1, raid10, raid0+1.

There is no indication that a raid is to blame if he managed CD's to go bad. More likely is operator error or corrupted data on the primary disk from the start before the final failure.

Not true at all. CDs go bad whatever you do (http://en.wikipedia.org/wiki/Bit_rot). It is unrelated to the RAID issue.

Well here's the skinny,

After repeated room temperature tries with no recognition I left it in a cold room overnight. Tried it the next day with the HD at 45 deg F. All the files and folders appeard on the drive.

I copied everything onto the new drive. It shows to be identical file sizes etc. But everything is corrupted.

I tried Ghost and the files were the same. So I checked everything from the new drive in Vegas and Windows Media Player. No help. Then I tried the same from the bad drive; same result. I was surprised the drive would still read after several hours of goofing with it.

So I rebooted and tried again. Same thing.

I disconnected every drive except C and the bad one E, no change. I relocated E to other SATA ports on the MOBO and it still came out the same.

At this point I thought I was scr***d. I decided to give up on the drive and data and reformat the drive a play with it with archived data. Then I discovered the most important of my files were copied on to My Documents in C.... WOW! Who did that? I don't remember doing that but WTF, lucky me.

So I've backed up the precious files, reformatted the drive in question, and I will see how it acts just for the experience.

About that freezing thing. That sounded kinda scary which is why I tried the cold room first. Maybe I should have cooled it down even more.

Here in Houston, Texas the weather changes rapidly from hot/warm humid today and cold dry the next.

My computer is inside a well insulated building but I do not leave the AC on all the time. Sometimes a few days pass before I turn on the Computer. There is probably as much as 30 degrees variation in temperature inside the office and humidity in Houston varies from a normal high of 80-85% to less frequent lows of 40-50%. Probably a factor.

I'll stop boring you guys,

thanks everyone for your input and three cheers for the knowlege base here at DV Info Net


Thomas Edward Bufkin

I ran a 24/7 database suite and thanked my lucky stars many times for RAID5 or mirroring to get thru surges and other short term disasters. However, for longer term protection and especially self inflicted "disasters", you need a SIMPLE backup medium, one that is NOT on line all the time. Thru 1999, that was always tape. Here we copied the live dataset to a new tape or to one that was, say, 10 generations old and rotated them. One went home with the sysadmin, the other 9 were in the safe. We ALWAYS bought tape drives in pairs: If the live machine melts, you will need a drive and software like the original.

Translating that to a world where discs are so cheap. I am determined to keep "My Archive Video" of .m2t and .iso files on my "C" array. Every 2 to 4 years I'll build a new machine with another 4 disc RAID5, I edit to a work volume on another pair (ie #5&6) striped RAID0. I then only use my eSATA external to put new projects on. As such it only runs once or twice a month. Occasionally I plug that eSATA drive into my Blu-Ray notebook and pull .m2t or .iso off it to check all is OK.

Edit: in summary, if it is your only copy, its not a backup!