Hello, doing the obvious, fixing bad blocks. But times have changed. Drives are so big that scans with little feedback take days, weeks, and one of my drives is still scanning for over a month now. So trying new scanners that do quick overview then let me target only potential badspots. Salvation looked promising, but only finds the harddrive on my laptop, and cant deal with the rest of my PCs. Now trying MHDD. I target a bad area and set it to loop mode. First erase mode, then map mode. But the bad area never vanishes. Every testing loop reexamines everything all over again. The damaged area is never going to come back so I want it to mark it permanantly gone so the drive eventually scans fast and smoothly. Will MHDD do this, or do I need to try another software?

You need special hw+sw combo tools and knowledge to deal with internal drive selfscan. You're wasting time, no software in the world will help you. Or ask a pro, it doesn't cost the world. About the tools, from about 500$ brand-specific, to several thousand $.

Bummer, I havent tried "makebad" yet but it sounds like a utillity to do one block at a time instead of a scanner to make them all bad if over 100mSec. If a proggy did that, then a crashed drive could be made very relyable again. I would be glad to dump 20% of a drive to get relyability when bad luck strikes. Waiting no more than 100mSec for a block would also drastically speed up the completion time of large drive scans.Futher making it worth the capacity sacrifice.I got one drive been scanning so long, the bearings will be worn out before it gets all the bad blocks marked. And its only down 3% on capacity after a month of bad block after bad block. And most likely, a lot of borderline blocks are getting by that will be nothing but trouble because I cant change the wait time threashold. Any more its cheaper to dump the drive for a new one but it costs me a slightly higher electric bill to keep them on old pc's running scandisk.

If you say "makebad" will solve... you haven't read the manual... no comment. Anyway you don't need any help, you know everything. Fine.
You didn't mention the brand and the model of the drive.
"Any more its cheaper to dump the drive for a new one but it costs me a slightly higher electric bill to keep them on old pc's running scandisk." : make up your mind... throw away or want to have it repaired ?
To end this thread : no one can help you, no software, no emule, no google.
Your drive (still unknown brand or model...) has maybe a weak head (dying) or need to be serviced (maybe FW or HW problem to be diagnosed). I know the hope has to be the last to die, but when there's no homegrown solution, the answer is NO. Anyone can post his price for diagnose and maybe a selfscan attempt ?

Quick method, quicker than the rest. Partition either side of the bad blocks. Leaving the bad blocks alone with no partition made for them, problem fixed.

opps, it appears somebody found what I said offensive. Oh well, it happens but wasnt intended. I have not found a manual for MHDD and the F1 and MAN descriptions are short and cryptic, so I will have to play with this to understand the full capability. I have done some hunting around the forum to get more insite and found quite a few cat and mouse chases others engaged in trying to figure out the features. Kind of a bummer but the HVAC forum is much much worse as they really try to keep AC repair details within the professional comunity. A lot of members over there get in really heated debates and thats when they start spilling juicy details. I know enough now to never need to call an AC repairman so its all good.

But as to Harddrives, taking them apart to fix them is such a tedious crap shoot, I hope I never mess up on backups to have to travel that path. But walk away scan tools to recover in a pinch is certainly worth while.

I already DLed a Makebad proggy but I just discovered that it is also in MHDD. I also discovered about the .cfg file and changed it to see what happens. I get more W's now on an erasewaits scan so maybe this can really cleanup a bad hardrive. But its not clear if a format of somekind can remove these markings. Its also not clear how much remapping can be done before I run out of remapping capability, or how permanent that procedure is as well. If remapping space runs out, will it tell me, or will it simply start reusing the space? These big drives take so long to scan that we have to know what exacly is happening because its not practical to go in and check the whole thing out to be comfortable in the relyability. Thats why I will hope erasing wait blocks over 100mSec is format proof.

Bnice has a great idea. I have one drive that has a small area smack in the middle. When Erase hits it, the head goes click and blue screen comes up and says the drive is toast. But starting erase a ways on the otherside, and its all good again. It looks like I will be making a relyable 2 partiion drive to get around that bad spot.

Oops, I was going to add more with edit but it says I cant, so Ill just continue here...

I have DLed emule stuff for years so I have lots and lots of drives. A Seagate 300Gig has thousands of UNCs in roughly one area that I hope to clean up but only a tiny fraction of the capability over all. Another Seagate 400G has been in scandisk over a month at the 85% level. No warning on its failure so maybe a head quit on that one. Eventually I will try MHDD on that one to get more details about what is going on. But no doubt I will eventually dump the warranty ones for a replacement when I am done. I have several WD2000jb's with flaky blocks I will try erase waits on as well as several of those WD1200jb's that need attention. The newest member of the distressed drive family is a Seagate 7200.11 1Tb drive. I just got that sucker loaded up 3 weeks ago, and now it fails to be noticed in the bios. With no drive on the controller, booting zips right thru, but with that drive connected, it hangs for a bit then finally says no drive connected. Its no longer seen on any of my PC's and there was no warning or impact. Heat is not an issue as its well cooled. So I guess maybe a case of the FW forgeting itself? Oh, it spins up and the head sounds like its getting ready to play like normal. Nothing sounds unusual compared to the other seagates, just not there for the bios. So watch out folks. If you bet the farm on these 1TB drives, and not have 1TB of backups, I know you cant help crying, but you will have asked for it.

Don't worry, if a drive fails we know what to do. @bnice, have you heard about error growth? Timeout for me.

Perhaps you annoyed BlackST with the title of this thread... Is it 'Intelligent' to try to store data on a dying drive?

No he didn't annoy me. Only all we pros here deserve to be respected some more. I am a service engineer not a book writer, a problem should be explained in 4 lines or less , unless it's a very technical one.

How things should work :

POST :

I have a drive with bad blocks, what should I do with free tools ? Can you recommend a quick way to try a fix ? I am trying with MHDD but I don't know anything about this tool (optional : I don't want so spend 1 cent to have this fixed and I don't care about the data or reliability)

ALTERNATIVE :

I did not understand some commands of MHDD is there anybody can give me 2 min. of his time explaining XXX or YYY ? Thanks

ALTERNATIVE #2 :

How can I fix a drive with bads or if not fixable by myself how much does it cost ?

REPLY :

Try MHDD with ERASE command on areas involved then SCAN again (youì'll lose the data ERASEd). If the bads are gone 75% the problem is fixed, anyway why did the bads came out? Stilla surface problem could arise.There's a manual for MHDD available. Otherwise, it's a surface problem and can't be fully fixed with any free tool and it's beyond you.

NETIQUETTE is not an optional: a clear, polite and eligible for replying post (maybe just for me) would only take 1 to 3 lines (approx. time to read : 20 seconds) and the reply just 4 lines.
Right ? Almost all people here know.

There's only the power of knowledge and the strength of honesty. Oh, sorry, Maybe there was. And also I forgot humbleness...

Well, I dont want to burry a hard drive post in completely unrelated nonsence, but let me quickly clear the air. I dont think intelegent scanning takes months, so I would label tools that work in hours to be much more inteligent than what scandisk has been doing. Those are the tools that will atract my interest. Second, I dont know anybody here and vise versa, so let me assure you all that I have the utmost respect for professionals that post resonably professionally. I am pretty informal and I care about knowledge and friendlyness. I couldnt care less about punctuation, social cast, race, or how big your head is.

You want respect, well so do I. Dont call me useless names and I wont call them either. lets just talk friendly like so we all can discover something we didnt know an hour ago. The story of my life doesnt need to be told to have to earn respect. I am a human being and that is all thats required. But if it makes a difference, I am a Computer engineer, working freelance as factory machine build and maintenance. I make tons of money, fixing what some would call the most impossible fixes to get down factory machines running again. I design and build the electrical end of special machines that do whatever needs done. In my spare time, I DL tons of movies on Emule, I also hand build Lamborghini Diablo cars much cheaper than Lambo sells them for. I have also built houses and decks, but I cant seem to figure out how to paint what people would call art with an airbrush. So there you go. But this is the internet and I could also be a complete BSer, so we are back to the human being part of all this.

So lets just talk friendly technical details and experiances and leave it at that. Everybody will be much happier as that is what this forum is suposed to be all about.

Nuff Said.

The growth-list will be screwed, as his drives are now. Not much he can do for free, with no tools for diy.

Get the data of the failed drive(s) use a sector imaging tool. Working from the image file to copy some or all of the recovered data onto a new drive. You may not get back all of your data. The longer you play with the drives, the more data you will be losing. Why worry about repairing this and that sector. As you can replace your free emule data, have fun.

Hmm, this mhdd isnt too bad as I play with it more. I completed several drives now that had a manageable quantity of bad blocks. Probably a tiny little head scrape. Did the erase, then remap, then erasewaites 150mSec. Ill use these for emule temps that get heavily exorcised to see how they hold up. The other drives with millions of BBs however, are seeing the same problem as scandisk. This thing spends 10seconds on every hopeless block making repair completion unrealistic even if its a small fraction of capacity. I will settle for probing with Mhdd for the damage boundaries and partition around it. Eventually, all warranty drives will be RMAed.

But that 1T byte drive has my curiosity peaked. I saw this Utube video about live swapping and I like that. The drive doesnt act like any other damaged drive so maybe the platter software is just corrupt. I got the drive a few weeks ago, so Ill see if bestbuy will have another close to the same manufacturing run. I am not inclined to open it up with the micro densities making dust look like boulders, but swapping boards that are live is really up my alley. Ill post on my experiance and if you have any tips you think I might need, feel free as eventually somebody else surely will be pinched by a big drive failure.

It sounds like you've got way too many bad drives, and way too much time on your hands. If you're considering that opening a drive will make it somehow more reliable, I won't argue with you. You seem like you've got a real good handle on things, so I'll bow out now.

Next stop, Freezer!

There are 2 stop before the freezer - head swap and platter swap...

I went and counted them and I own 116 5-1/4inch drives, and 19 laptop drives. Thats roughly 24Tbytes of storage. I am kinda addicted to emule. One laptop drive is click of death, and 8 of the 5 1/4s are in various states of distress. Awhile back I bought a 400gigy seagate and out of box it was clearly a 160giger even down to the pcb, with a 400 gig label stuck on it. I imediately took it back to best buy and they were amazed but swapped it out for a real 400. Once when I had 21 drives lined up like dominoes serving up files, I slipped on a clear plastic wrapper on carpet and landed on the drives. I was hurting, and the squalling and squealing was absolutely unreal. 8 drives were destroyed and 5 more recovered with lots of BBs. I eventually RMAed all the damaged ones useing the names of all my neighbors so WD wouldnt have a cow. Another time I blew all 5 drives in a cage when I accidently connected the power plug backwards. It was right at bootup and I just remembered to plug them in after I hit the power button and tried to beat the bios. Yes the plug is polarized, but the pins did touch. These were all seagates, and my neighbors again helped me out. So yes, I have had my share of HD bad luck, so I am very big on backups. I have 4 3ware raid cards but non of the storage is raided. Just make them JBOD so one pci slot is used. I use every letter of the alphabet and my room is trully a computer room. I used to have 7 21" trinitrons in there when I daytraded stocks, but I finally hauled them out to make room with flat panels and a Belkin 8KVM. Things are much nicer now and not nearly as warm. In total, I have built things up to 2 8u PCs, 5 towers, and 5 laptops all ready for action. My Lambo kitcar is going slowely but Ill get back to it in the fall when I lighten up on the Emule.

Well, getting back to drives, its no doubt that surgery is not relyability and is only a way of possibly recovering files. And I dont think I would really open one up, but who knows what I do from day to day. Right this minute Im not inclined, but I only mess with these drives for practice and curiosity. You never know if the practice may come in handy some day. But I am not inclined to tear the warranty stickers off for RMA reasons, so I will just stick with PCB practice and temperature experiments to see what happens. I may not get anywhere but its pretty cool just trying.

do something

Alright alright, I did something. I went to best buy and got another 1TB seagate just like the dead one I got. I picked thru the stack and got the one as close as possible to the serial number of the one I have. Dead one is 9qj0twff. New one is 9qj0wwwn. As it turns out, they both have FW sd15, and site code kratsg. On power up, they both sound the same in the noise the head makes, but the new one does a few more clickety clicks at the very end before they both go silent untill needed. I start winxp on the new drive, partition and format it, then shut winxp down.

Now the experiment. On the good drive, I remove all but one screw by the head and motor connection so I dont have to fumble with too many screws. On the dead drive, I completely remove the board and put the drive next to the new one at the ready. I start Winxp and go into control panel and sleep drives after 3 minutes. When drive spins down, I unscrew last 2 screws and lift pcb as I go, so screws stay in the pcb holes. I lift board off and set it down on bad drive and screw it down. Sleeping board is now on bad drive and everything went smoothly. I then bring up explorer and click on the second drive. The dead drive spins up and then does the click click, click of death sound, over and over again. Explorer eventually gives up and the drive vanishes from the tree. I give up and power things down. I swap the boards back, and power them back up. The new drive seems to work great and the bad one goes to sounding almost normal in head activity but is invisible to the bios.

Oh well, it didnt take me long to do all this but its clear that live board swapping doesnt fix everything. I am not going to tear up the label and open the drive. But this thing is huge and the data is so dense, that microscopic differences in head alignment mean that no 2 head stacks are alike. So I cant see head swapping as being much use even if you were in a desperate situation. Same goes for plater swapping. Once these are formated at the factory with the track location data, only these heads will read it and if they fail, the drive data is a total lose. Thats what I am thinking. So feel free to voice your opinion if you think I am wrong about this.

I think what they are trying to say is this.....

Fine if you want to try to find out why a drive doesn't work, some of the people here are hobbyists....like myself.

No one in their right mind trusts a faulty drive with important data.

Drives with many bads have them for a reason, PCB faults, power supply faults, weak head, bad media, SA problems.

When the G-list is full the drive is toast anyways. Unless you have about 12K in tools : )

to fix to learn is OK if you ask nicely this forum can be of assistance.

Sometimes it's not worth it for $85.00 for a 300GB hard drive.

Fastest way to re-check a drive. WDclear to write 0's to entire drive, & factory format tool, or MHDD and set the time for time out, then factory format software.

Anything more than about 4 hours for a drive is a waste unless someone is paying to get data back.

Spending three weeks scanning a drives surfaces is guaranteed to kill it

_________________
I think ST is a sub company of Lucas