I had a lot of data backed up on an external hard drive and I recently tried to access it and it was all gone(OS couldn't access HD). Seems the MBR got corrupt or possibly the partition table. It got me questioning why it went bad. It was not used or abused in the time it was working and wasn't(probably 3-5 months).
Only thing google turned up was
formatting link
It mentions that HD's automatically refresh(I guess sorta like dram?) the magnetics when they are plugged in. Is this true? It could explain why the external drive went bad and will help keep it from happening in the future.
So I'll I have to do is power it up and the drive's internal mechanics will refresh all the data with any need for me to initiate it? (e.g., copy the files, view them, etc...)
think about the nowadays HD's with their large capacity. To refresh the complete data stored on a half filled HD would take hours. It couldn't be true.
"Refresh"?? Idiot does not know his backside hole from a gopher hole.. What you can do that is not expensive, is BUY a copy of Spinrite. It installs / runs from a MS/PC DOS floppy and supports all of the HD formats including Linux. Use it on a semi-regular basis to check the integrity of your hard drives; IT will refresh and even recover (to a certain degree) weak / bad data.
Slightly related, I was at Walgreens, restocking the engineering chocolates bin, when I say a sign offering media data recovery services. In a drug store! I didn't get the details.
Walgreens is surprisingly high-tech: I think it was Mike Terrel? who was recently mentioning that they have a fancy machine that refills inkjet cartridges as well!
I bought a $10 shortwave radio from Walgreens once, just as a novelty. The performance was -- not surprisingly -- pretty lousy. :-)
Well, as I understand it (from talking with several hard drive engineers in the context of my job) he's somewhat correct.
With the increasing areal density of hard drives, the size of the magnetic domains on the platter (i.e. the individual "bits") has been decreasing, and the tracks have been packed closer and closer together. This leads to several significant effects:
- The amount of magnetic energy stored in each bit "domain" is being progressively reduced. It's approaching the point of becoming equal to the random thermal energy in the material, at room temperature... which means that the bits can become subject to thermal self-erasure.
- The magnetic field from the write head, as it writes a track, overlaps to some extent onto adjacent tracks, and can weaken the stored magnetic field in those tracks... in effect, partially erasing them. After a track is written enough times, the fields stored in the adjacent tracks can weaken enough to cause them to become unreadable.
- Even the tiniest speck of debris (e.g. dust, tobacco smoke particles, flakes of material knocked off of the platter by the head due to impact) can temporarily or permanently obscure the data in a bunch of bits. For this reason, modern hard drives incorporate quite a bit of preventive error-detection-and-correction maintenance into their firmware, so that they can detect tracks/sectors which are "going soft", and rewrite the data before it becomes too scrambled for the drive controller to read. Hard drive data is written with embedded error correction coding (typically Reed-Solomon), so that a certain number of bits in each sector can be correctly recovered if they aren't readable. The drive controllers will detect situations in which a sector's error rate is approaching the danger threshold, and will rewrite the data (or move it to an alternate sector in a way not visible to the host computer) when this happens.
Some of these firmware-based error corrections are done automatically, while the drive is in normal operation or is sitting idle. Others can be initiated from the computer, typically by issuing a "S.M.A.R.T." command to the drive (this is probably what modern versions of Spinrite are doing on IDE and SATA drives).
What I consider a good thing to do, is:
- Use S.M.A.R.T.-enabled software (e.g. SpinRite on Windows, or "smartctl" or a similar utility on Linux) to enable "automatic off-line data collection" on the drive. This tells the drive controller that it may automatically do an occasional read-scan of the hard drive, when the drive is otherwise idle. This should enable the controller to locate any sectors which are becoming hard to read, before they "fade out" enough to overwhelm the Reed/Solomon error correction coding (at which point the data in the whole sector is usually lost).
- Occasionally perform some of the S.M.A.R.T. on-line self-tests... the "short" test checks basic functionality, the "conveyance" test checks for forms of damage which can occur during shipping, and the "extended" test does a full surface scan.
- Keep an eye on the S.M.A.R.T. warnings... if the drive controller raises an alarm for any reason (e.g. bad sectors), do a full backup of the drive onto another disk *immediately*.
- Doing this on drives that you use infrequently (e.g. off-line backup or archive disks) is good practice. If you have permanently enabled the S.M.A.R.T. self-checking feature, it may be enough to plug in the drive and turn on the power... connecting it to a PC may not even be required, unless you want to check the results of the automatic self-scan. Keeping a backup of critical information, on a completely different disk or CD/DVD/tape, in a completely different location, is always a good idea.
--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
In the early 1980's most/all Walgreens had computer terminals and label printers for prescriptions. Worked out quite well for refills etc.
Many of the higher volume pharmacies here, have not just machines for counting pills (those have been around for a while), but a little computerized assembly line that takes an empty bottle, fills it with the right number of one of any number of the more common pills, and then labels it. I would hope at some point a tech checks the computers work (at least for correct pill type).
How do you know it is any of these? Does the drive spin up? Various mechanical things can prevent the read heads lifting - a fault which can often be fixed by giving the drive a quick jolt or sending it back to the manufacturer via post where it will arrive working and be put in the no fault found bin. Even if the OS doesn't recognise the drive as formatted it should recognise the USB connection and drivers.
There are DIY recovery programs that will sort out most faults on a physical drive provided that the control electronics are still working. And there are expensive services that will recover stuff for you at a price.
Not on that time scale and losing a few bits would normally be correctable allowing diagnostics to warn you. You get silly access times when drive media is borderline functional. My neighbour had one where Windows took 15mins to boot and XL half an hour to load. Some helpful person had turned off SMART error detection for him so he didn't see the enormous number of warnings of read errors!
Sticktion is a much more likely cause of failure. It was very common at one time just after CFCs were banned.
I think you are barking up the wrong tree here. Much more likely that something trashed the MBR if it is in fact missing. If you know the drive model you should be able to find one on th emakers support site to refresh from - I have never used harddrive diagnostics on USB/Firewire connected drive so I don't know how well they work like that. You might need to put the drive inside a PC to work on it.
that won't fix the partition table, the master boot record is only useful for booting from the drive. (and even then only works if attached to a valid partition table)
To fix the partition table you need either a fdisk that doesn't trash the partition (not the MS product) and a memory of what it contained, a copy of the partition table, a partition table guessing tool like gpart, or an expert with a hex editor.
You can't fix the partition table without a good MBR. There are partion tools that will fix the rest. At one time I saw dead hard drives where some malware chaged the MBR. All the anti virus software said you had to reformat the drive to repair the MBR. They were wrong.
--
You can't fix stupid. You can't even put a band-aid on it, because it's
Teflon coated.
It seems Larry and Jordan are apple experts.This explains all you need to know. ;-)
It wouldn't hurt to run chkdsk once in a while.
We could learn a lot from that NSA MAID (massive array of idle drives) regarding data retention. Or perhaps NARA or LOC has tips. Most of what I see on the internet regarding data retention tends to be old geek's tales.
Figured as much. When the OS detects that there may be volume corruption, a chkdsk is scheduled automatically on the next boot. Oh well, at least Glimmermoan has dropped the thread now that he's realized he was wrong (yet) again.
Probably because you don't know how. Why don't you just admit you were wrong in that chkdsk is also exist for NTFS partitions?
Have a friend (if you even *have* any) take them all to the roof of a twenty story office building. You stand below and have him drop them on your head one-by-one. See if that knocks any sense into you.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.