Page 2 of 2 FirstFirst 12
Results 11 to 20 of 20

Thread: ******HDD odd********

  1. #11
    Now it's working..............I still stand by my hardware failure scenario........."I think this HDD just died, but it worked with linux." Not only that but I'm betting I could act out that scenario on an old HDD. If you can prove it to be flawed that's great!

    I didn't see any error messages it just went on idle. This is an authentic OEM CD..... it works just great without an Internet connection during install......except on that HP HDD. It worked fine on others before this HDD and as I confirmed for myself last night, it works fine on a different HDD during install without an Internet**middleman**. I just thought I'd throw it out here to hear the opinions of our professionals on this board.


    thanks, nihil,foxy

  2. #12
    The Doctor Und3ertak3r's Avatar
    Join Date
    Apr 2002
    Posts
    2,744
    During a WinXP install there was an option (pre sp1) to d/l latest drivers.. this only showed when you had an active network connection.. I havent seen it in the later versions of XP..but then I have a policy now of not connecting the LAN untill after the install.. Why this policy? it has nothing to do with privacy.. I did an install live on a customers network.. and managed to create a problem in 3 systems where the solution was to replace the NIC's ..

    It is an interesting situation you have there.. is it possable to run some testting tools over that HDD.. any that give you the SMART status.. ..do you have or can you get a copy of HDD Regenerator..(found also on the Hirens Boot disk - I dont know how Legit this one is) this prog does a read write test of the entire HDD (it is slow- a 20G hdd will take about3hrs).

    While I agree with CHSH.. hardware is hardware.. and it either works or it dosent.. unfortunatly certain software calls place hardware at the limits of its capable operation.. and any timing issues (like in memory) will cause errors.. isssues where a data words are read from two memory slots .. I am not sure of what the HDD problems could be except perhaps a Controller issue..
    The tests I mention earlier gives a general idea of the health .. I suspect the Media is ok..
    How Linux and Widows use the hardware.. the built in timings.. etc will cause different problems on the same hardware.
    So I take it that:
    even without the internet connection capability ..
    and that one HDD, the install still just stops.. no errors.. just goes to sleep?
    yet same machine and a different HDD all is OK?
    What happens with that HDD in a different system.. ?

    I am intrigued as to the cause of the problem
    "Consumer technology now exceeds the average persons ability to comprehend how to use it..give up hope of them being able to understand how it works." - Me http://www.cybercrypt.co.nr

  3. #13
    Originally posted here by chsh
    Hardware is hardware, if it works, it will work. If other software is working fine on a given set of hardware, start working towards it being a software related issue.
    I assume your statement is anecdotal.
    Either you're over simplifying all hardware, or you know a lot of the unknown and should work for NASA’s mechanical engineering branch! Because you've patently found a way to throw out, let's say, a trillion different variables! Does a battery have to be at 100% to work? Did it ever occur to you that not all of the components that make up certain hardware’s have to work for the hardware’s to work? A lot of hardware in this world can work at below 100%. I suspect HDD's not to be the exception here. A PCI slot can go bad on a motherboard and the board can still be operational.

    Is it possible that the arm inside an HDD that holds the read and write heads could try and make a pass over a bad sector, track, or cluster on a platter and idle during an WindowsXP install, but not during a *nix install because the head never reached the bad sector, track, or cluster on the platter? (Because it didn't need to go that far for the *nix install)

    This is one of a trillion different, possible variables I'm asking about?
    "things will go wrong in any given situation"

    chsh are you going to say it's not possible?


    She's still getting the new HDD. I found something out though, apparently there's a hidden partition on this drive. Some tools bewrayed some type of a large sum of unmovable sectors on this drive. Is that possible to move, delete, or corrupt? Does Hewlett Packer have some type of special HDD's made from the HDD manufacture? I'm thinking it has a part of the platter that the head can't reach to write or it has a built in chip on the HDD mini-board that prohibits writing in that area of the platter.

    What do you guy's think?

  4. #14
    Senior Member
    Join Date
    Oct 2001
    Posts
    786
    Most HDD's start writing from the outter-section of the disk. This means it goes outter->inner. For a given angular velocity (say, 5400 RPMs) the outside of the disk will travel faster than the inside due to its radius, and if the data is organized correctly this means the outter (beginning) sectors will have the highest transfer speeds. At the very least CD's & DVD's work this way.


    Some time ago there was discussion about hidden partitions. IIRC some companies did this to help their recovery CD's -- I don't know if HP was one of the companies but they probably were. It sounds like the disk is from that era, but I have no idea if the stories of this happening are true...today it is mainly a "trick" to corrupt the FAT to make your 120GB disk look like a 200GB disk, although it leads to a lot of data corruption on the disk... (supposedly back in the day the FAT tricks just allocated the hidden sectors -- though today if there are any hidden sectors is for error recovery to swap bad sectors with good ones during access...)

    WindowsXP installs can hang for all sorts of reasons. Try to take everything unnecessary out and then install. If that doesn't work, hopefully the second 20GB disk you've purchased will work out alright...

  5. #15
    Most HDD's start writing from the outter-section of the disk
    No you're right, I read this the other night. http://img56.echo.cx/img56/9018/111113vo.jpg
    My scenario should work vice versa though, no mater which way or how it writes.

    Try to take everything unnecessary out and then install.
    Oh, I have her working, she works just fine now. All of the sudden she just wanted to talk to the Internet. Now that's the anomaly I don't understand! I have Xp on it now, and I'm keeping her for further research....... I'm leaning toward this hidden partition stuff as the potential culprit.

    I still would like to have some insight on my scenario.


    thanks tim

  6. #16
    Senior Member nihil's Avatar
    Join Date
    Jul 2003
    Location
    United Kingdom: Bridlington
    Posts
    17,188
    Hi !mitationRust

    Yes, several of the brand name computer manufacturers use hidden partitions, and HP were at the forefront of this. In fact, if you buy a modern HP box you don't even get a recovery disk, it is all on a hidden drive.

    These partitions generally contain diagnostic tools and recovery data. In the old days these would come on CDs and 3.5" floppies, but manufacturers discovered that people tend to lose these, so they went for the hidden drive/partition to cut their helpdesk costs.

    Now, you have an HP drive.............well that is a clue because HP don't actually make HDDs. If it says HP, it has been made by someone else and badged with the HP logo. Obviously it will have been made to HP specifications. The hidden partition will probably contain the user manual and diagnostic/recovery tools for the original HP box that it came out of.

    I do not think that the hidden partition will cause the problems you describe. After all, you could not access it even after formatting and using autoclave. I suspect that it is protected by an EEPROM chip on the drive itself. Anyway, it loaded XP and worked in the old machine?

    Bad sectors are a possibility, and you should be able to find these with a surface scan, as Und3ertak3r has suggested. I am slightly surprised that these were not detected and repaired (isolated) as part of the installation process, and that you did not get an error message
    at least from autoclave, which I presume that you ran over the whole disk?

    Of course it could be that the HDD is getting more damaged each time it is used, so the freshly damaged sectors have not been isolated?

    What really confuses me is the attempted network connection. Particularly as I would expect that most OEM builders configure and test the system BEFORE attempting a network connection.

    I would like to know the results of a surface scan. If you get any bad sectors reported, run it a second time and see if you get any more............that will tell us if the HDD is at death's door

  7. #17
    Senior Member
    Join Date
    Nov 2001
    Posts
    1,255
    Originally posted here by !mitationRust
    I assume your statement is anecdotal.
    Either you're over simplifying all hardware, or you know a lot of the unknown and should work for NASA’s mechanical engineering branch! Because you've patently found a way to throw out, let's say, a trillion different variables! Does a battery have to be at 100% to work? Did it ever occur to you that not all of the components that make up certain hardware’s have to work for the hardware’s to work? A lot of hardware in this world can work at below 100%. I suspect HDD's not to be the exception here. A PCI slot can go bad on a motherboard and the board can still be operational.
    In each of the scenarios you mention, one particular piece of hardware is failing. I don't consider a battery to be "hardware" in the context of this conversation, but yes, if it is working, it should remain working. Providing power is not an aspect of a battery functioning properly, it is an aspect of whether the battery is charged. A drained battery still works (talking rechargeables here). A PCI slot going bad on a motherboard means that the PCI slot went bad. The hardware in this case is the PCI slot, whether it be connectors, improper power regulation, or what, it's still the slot itself that has the problem. I have NEVER encountered ANY hardware that just plain worked in one OS and not another BECAUSE of the hardware. In one way or another it was ALWAYS the software.

    Is it possible that the arm inside an HDD that holds the read and write heads could try and make a pass over a bad sector, track, or cluster on a platter and idle during an WindowsXP install, but not during a *nix install because the head never reached the bad sector, track, or cluster on the platter? (Because it didn't need to go that far for the *nix install)
    Test your theory out then, see if it holds water. Run something like a dd if=/dev/urandom of=/dev/hd?. Fill the drive, see if you start getting write errors. Oh wait, you said you got it working by connecting it to the internet...

    This is one of a trillion different, possible variables I'm asking about?
    "things will go wrong in any given situation"

    chsh are you going to say it's not possible?
    I would bother answering this, but I mean, as far as I'm concerned the discussion is over, it WASN'T the drive, arguing about what it maybe might have been if a set of conditions that aren't true were true is a waste of time.


    She's still getting the new HDD.
    That's nice of you.

    I found something out though, apparently there's a hidden partition on this drive. Some tools bewrayed some type of a large sum of unmovable sectors on this drive. Is that possible to move, delete, or corrupt? Does Hewlett Packer have some type of special HDD's made from the HDD manufacture? I'm thinking it has a part of the platter that the head can't reach to write or it has a built in chip on the HDD mini-board that prohibits writing in that area of the platter.

    What do you guy's think?
    This might surprise you, but almost every SCSI hard disk ever manufactured has "hidden sectors". On SCSI drives, it's usually ~9% of the drive space dedicated to bad sector recovery. This is why SCSI drives are generally much more stable and durable, and why an 80GB SCSI drive is actually billed as 73GB (IIRC). It's quite common by OEMs too. Dell puts on small partitions (which you can actually see). Compaq used to put invisible to everything partitions that were used with compaq-specific utilities and would only work if it was in an actual compaq box. It wouldn't surprise me if it was such a system on an HP machine.

    Originally posted here by Und3ertak3r
    While I agree with CHSH.. hardware is hardware.. and it either works or it dosent.. unfortunatly certain software calls place hardware at the limits of its capable operation.. and any timing issues (like in memory) will cause errors.. isssues where a data words are read from two memory slots..
    Timing issues are induced when you have incorrectly configured software at an extremely low level controlling the hardware. I think you'll find most of the software-induced hardware failures will end up being very low level.

    I am not sure of what the HDD problems could be except perhaps a Controller issue..
    The tests I mention earlier gives a general idea of the health .. I suspect the Media is ok..
    How Linux and Widows use the hardware.. the built in timings.. etc will cause different problems on the same hardware.
    I'd really like to see a situation where Linux and Windows use a piece of hardware and it causes a hardware problem that does NOT relate to drivers.

    EDIT: Let me put it this way: If I am using a tool with a duty cycle (say a Soldering Iron), and I choose to ignore the duty cycle, and the tool breaks, is that the tool's fault? Fundamentally, it's the same thing here. The low level software (me) is controlling the hardware and telling it what to do, and the hardware is happily doing what it's told.

    Take it to a higher level. Say I'm using a Soldering Iron, and I think I know which end is the handle, but I am wrong. You think it's going to work at all?
    Chris Shepherd
    The Nelson-Shepherd cutoff: The point at which you realise someone is an idiot while trying to help them.
    \"Well as far as the spelling, I speak fluently both your native languages. Do you even can try spell mine ?\" -- Failed Insult
    Is your whole family retarded, or did they just catch it from you?

  8. #18
    Senior Member nihil's Avatar
    Join Date
    Jul 2003
    Location
    United Kingdom: Bridlington
    Posts
    17,188
    I'd really like to see a situation where Linux and Windows use a piece of hardware and it causes a hardware problem that does NOT relate to drivers.
    That is not what is being said. The suggestion was that as the linux distro takes up less HDD space than XP, could it be that linux worked because it never came to a possible bad sector that XP tried to use.

    Test your theory out then, see if it holds water. Run something like a dd if=/dev/urandom of=/dev/hd?. Fill the drive, see if you start getting write errors
    Yes...............a surface/diagnostic scan as already suggested by undies and myself, should also reveal that. However:

    !mitationRust said that he had run a full autoclave on the drive............that software overwrites the HDD several times, so he should have seen write errors then. Also I would have expected XP to produce errors if it encountered bad sectors.

    The really weird thing is the attempt at a network connection as far as I am concerned.

    With HDDs, they will frequently work even though partially damaged.............here the damage is to the platter, so the drive will function with the remaining good sectors. It is really no different from a dodgy floppy.

    It may surprise you to learn that virtually no HDD ships in perfect condition. There are always minor surface imperfections. Its just that the OEM manufacturers "clean them up" They still work just fine though.

  9. #19
    AFLAAACKKK!!
    Join Date
    Apr 2004
    Posts
    1,066
    This is all really quite simple... you guys have done a really great job of overthinking all of this... While the fact that it had to connect to the internet is very strange... what the installer did is not... So it was trying to connect to the internet... it couldn't so it just kept trying to connect... therefore it's going to hang... it's as simple as that... I guess you should have gotten some kind of error like... "could not connect to the internet to download necessary updates" or something like that, but that's not important... What we all should be trying to figure out is WHY it had to connect to the internet, as it seems no one is really sure why...
    I am the uber duck!!1
    Proxy Tools

  10. #20
    AO Curmudgeon rcgreen's Avatar
    Join Date
    Nov 2001
    Posts
    2,716
    Don't confuse a hidden partition with the hidden sectors used for bad sector recovery.
    The hidden sectors you can safely ignore. That's a private conversation between
    the controller and the drive, and is not software accessible. It is not an issue.

    If there is a hidden partition, such as the kind used by COMPAQ, it will be defined
    in the partition table, and will be deleted when you fdisk the drive. The installer
    does this at the stage where it asks you if you "want to use the whole drive"
    or some such question.

    If you delete the hidden partition on a COMPAQ, you can't get into the BIOS
    setup, but I take it this isn't even the original motherboard, right?
    So that isn't an issue either.

    You probably didn't start this whole thing without scandisk, or some other test
    of the media, did you? Then it isn't a bad sector.

    There is most likely nothing wrong with the drive, since you successfully
    installed another OS on it.

    You have probably discovered a rare incompatibility between the drive/controller
    and the XP installer. Maybe the installer wants to connect to the net to
    get the driver to solve this incompatibility.

    Maybe it is demon possessed.
    I came in to the world with nothing. I still have most of it.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •