Who DOES NOT have a problem with corrupt images...

Discussion in 'Acronis True Image Product Line' started by hoser_d, Feb 5, 2005.

Thread Status:
Not open for further replies.
  1. brandis

    brandis Registered Member

    Joined:
    Feb 4, 2005
    Posts:
    60
    Hello,

    it seems that the answer is YES. Till now I did not find a corrupt Image on my external MAXTOR that I did copy from Server via NEC-USB20.
    I dont believe that it works because the read speed is flat in HD-TAch, this show only that its limited by the chipset. If the speed is limited by the disk you get slower results at the end of the disk (like in the other results I posted).

    Unfortunatly I did get many W2000 errors ID5 ID11 ID50 on the server, doing heavy work with the internal drives (copy or imaging). It seems something is not o.k. but I dont know what. Till now this errors does not corrupt the data.
    Perhaps I will change the IDE cable, or/and reduce the UDMA to a lower level to see if this helps.

    I did also some benchmarking on my brothers new IBM thinkpad R51 Labtop. Again comparing internal USB20 with internal 1394. USB20 dont seem to have problems on this pc, but again 1394 is FASTER and DONT USE SO MUCH CPU!
    BLUE is 1394 faster read speed faster burst speed only 5% CPU
    RED is USB20 slower AND 42% CPU utilization
     

    Attached Files:

  2. jmk94903

    jmk94903 Registered Member

    Joined:
    Jul 10, 2004
    Posts:
    3,329
    Location:
    San Rafael, CA
    Hi Stephanos,

    I'm glad to hear that your Server makes good backups and copies of images to the external USB 2 drive.

    When I said a "flat" response I should have said one that didn't have sharp drops in speed to 3-10Mbps from the maximum speed. You provided one HD-Tach report like that for the system that made corrupted images via USB 2 but was good with Firewire which didn't have the sharp drops.

    A gradual fall off in speed is clearly OK and due to the hard drive limiting the performance as you said. A completely flat response like my system is indeed due to the chipset or system limiting the speed to less than the drive can handle.

    I think the key is not to see sudden decreases in speed in the HD-Tach report, because it looks as though this leads to corruption.

    I hope a few other people will try HD-Tach and report their results for USB 2 external drives and even IDE drives and whether they see corruption or not.
     
  3. brandis

    brandis Registered Member

    Joined:
    Feb 4, 2005
    Posts:
    60
    Yes it would be nice to see this from all the users who complained corrupt images, but till now no one has done it :'(

    HD Tach does read the complete disc, you can see deteiled results in the EVENT.LOG file. But it dont write all the results in this file, it stops. The READ results are so many that they have to reduce it in some way to draw the graph (it would be nice to know how they reduce the data to draw the graph).
     
  4. I was having the image corruption problem, too. Acronis could not backup a good image. Even backing the c: drive up to itself, then manually copying the image to the d: drive did not work. MD5 showed the files as not identical.

    The c: drive and d: drive are two physical IDE drives on 1 IDE channel. No USB involved.

    Upon testing my memory, I found a memory problem. This was huge. MemTest86 1.51, test 7 was detecting a bad cell. I have since replaced the memory, and MemTest86 runs clean now. You can get MemTest86 1.51 here:

    http://www.softpedia.com/get/System/Benchmarks/Memtest86-plus.shtml

    You burn the ISO on a CD, and boot from this CD. You should run all tests (default) all night. It should NEVER throw an error. If it throws even one, you have a problem.

    Now I can backup my c: drive to my d: drive. Acronis reports no corruption when the image is checked. I have also manually copied files as large as 3.5 gigs and checked them with MD5. All now check out as identical.

    So for me, it was a memory problem.

    I will be getting an enclosure soon, so I will naturally test this. It will probably be a USB 2/External SATA enclosure. I want external SATA for the speed. I will test it in both modes and report back (could be a few weeks, though).
     
  5. jmk94903

    jmk94903 Registered Member

    Joined:
    Jul 10, 2004
    Posts:
    3,329
    Location:
    San Rafael, CA
    JDD,
    Thanks for this post. I suspect that memory problems are a major source of the file corruption errors, especially if it happens between internal IDE drives or partitions, but you have a proven case.

    I downloaded Memtest for a friend of mine who has infrequent, random corruption problems. I'll burn the CD and have him run it overnight or longer to see if he has memory problems also.

    I look forward to your future posts with the external drive.
     
  6. Tsu

    Tsu Registered Member

    Joined:
    Sep 16, 2004
    Posts:
    61
    Good for you JDD... finally someone has stopped the incessant CRC, MD5, Tach Test navel gazing and actually tried a fix. I'm glad you are back in business.

    From my Feb 7 post item #4:

    4 - download this MS memory checker, make and boot to the floppy and test memory - http://oca.microsoft.com/en/windiag.asp. You may just be using some memory addresses when doing the copy that are normally not acccessed during day-to-day use when DLL and other stuff get flushed out when not used. Although I quite like the registry hacks that keep system files in memory and not have them swap out to disk when not in use. Folks with fast drives will find that archaic. But are you also getting unexpected system reboots in the process of doing quite ordinary things? You may just have one bad memory chip. Remove it from you system and use what is left to try the LF copy.​
     
  7. metafizx

    metafizx Registered Member

    Joined:
    Apr 29, 2004
    Posts:
    35
    very scary thread.

    I just wanted to say that I have been using TI 7 for many image/restores for XP Home/Pro and W2k Server without any problems.

    I have created and restored from both PATA and USB 2.0 drives. Both Seagate and Maxtor external USB drives.

    Never had a failure yet. My images are definitely into the GB so they are not small.

    *puppy*
     
  8. brandis

    brandis Registered Member

    Joined:
    Feb 4, 2005
    Posts:
    60
    USB is not the only cause, but a very common

    Very nice testsoftware, thanks!!

    On my home PC the memory runs fine, USB20 not.

    I dont have any experiance with SATA.... but I would look to have IEEE1394
    as alternate, I didnt read of anyone reporting problems with it.
    With SATA it will be interesting to see if you can access this drive from the Acronis Bootdisk in the Linux mode
     
  9. Hoopster

    Hoopster Registered Member

    Joined:
    Feb 27, 2005
    Posts:
    25
    Corrupt Image on first use of True Image

    After reading glowing reviews of Acronis True Image, I purchased and downloaded the software this evening (version 8 build 796). My first experience has not been a good one. I guess I should have checked this forum before purchasing the software. I have had two HDD failures in the past 3 months, so, I needed a good backup solution. I purchased a Western Digital 250 GB drive which I placed in an external USB 2.0 enclosure primarily to backup my 120 GB and 250 GB Maxtor disks (and as a spare in case yet another Maxtor 250 GB disk fails).

    The image was created fine on my external USB disk; however, when I ckeck it, True Image reports it is corrupt. There are three partitions in the image (I created an image of the 120 GB disk as a test). I can assign drive letters and mount the images OK. However, only some of the files are accessible. I do not have a need to restore the image (and I hope I never do) and frankly, I would be reluctant to try it if the image is corrupt.

    My systems specs are as follows:

    Asus A7N8X Deluxe motherboard with on board USB 2.0
    NVidia nForce2 chipset
    AMD Athlon XP 2500+ processor
    Windows XP SP2
    1 Maxtor 120 GB hard drive (contains OS, Data, Applications
    1 Maxtor 250 GB hard drive (contains video and music files)
    1 Western Digital 250 GB hard drive in an external USB 2.0 enclosure

    Unfortunately, I see I have a lot of company in the corrupt image on an external USB 2.0 disk department. I have been reading these forums for an hour now. Is there any solution yet to this corrupt image issue? Obviously, it is unbelievable that backup software could create so many corrupt images.
     
  10. wdormann

    wdormann Registered Member

    Joined:
    Jun 27, 2004
    Posts:
    480
    As suggested previously in this thread...
    Get a multi-GB file on your HD. (TIB file, video capture, whatever...)
    Get an MD5 of that file.
    Copy to the USB drive
    Get an MD5 of that file.
    If they don't match, you can't possibly expect that it's the fault of TrueImage.
     
  11. Ed Y

    Ed Y Registered Member

    Joined:
    Oct 18, 2004
    Posts:
    86
    I never could pinpoint exactly what was causing my corrupt images on the external USB 2.0 drive I'm using but it's not TI. Once I started creating the images with a size of 700MB instead of default, I have NOT had a corrupt image problem yet and that's been in last 2-3 weeks, 5 or 6 images created.
     
  12. jmk94903

    jmk94903 Registered Member

    Joined:
    Jul 10, 2004
    Posts:
    3,329
    Location:
    San Rafael, CA
    After checking the MD5 values on a large file, greater than 1GB, copied to the USB 2 drive. If you get different results, file corruption is occuring with True Image not involved.

    RAM problems are another reason for corrupt files. Run memtest86+ overnight at least to confirm that this isn't the problem.
     
  13. David_O

    David_O Registered Member

    Joined:
    Feb 22, 2005
    Posts:
    10
    Re: Corrupt Image on first use of True Image

    Has it occurred to you that such errors might have nothing to do with Acronis and everything to do with your system? To that end, have you tried to write large files to your external drive using your OS alone and then compared the hash digests of the source and destination files? Until you do that, it would be a mistake to assume it is the Acronis software that is at fault. I spent the better part of my life troubleshooting electronic systems of all kinds, and problems like this are basic. You must root out the problem in a logical manner and not make assumptions. As a first step, I suggest that you zip up a bunch of files into a huge file on your internal hard drive, write that file to your external USB drive using your OS alone, and then compare the hash (MD5 etc) of the source and destination files. An alternative way to get a huge test file would be to have Acronis store an image onto your internal hard drive. Whichever way you choose to make the large file, do the test and compare the hashes. Chances are good that you will find that it is your system, and not the Acronis software that is at fault.

    Another potential gotcha are subtle memory errors. One can get along seemingly fine with marginal memory that tosses a bit every 200 MB or so of R/W activity. Apps will run just fine for the most part and so will your OS, perhaps with just an occasional crash or a strange result every now and then (which most folks would chalk up to buggy software). But try to copy a 1 GB file from one drive to another and, as should be no real surprise given the marginal memory, the destination file might well have few bits flipped in it.

    Just for the record, Acronis TI 8 writes perfect images to my external Seagate 300 GB USB 2.0 drive.
     
  14. Boskey

    Boskey Registered Member

    Joined:
    Dec 31, 2004
    Posts:
    5
    I have a couple of USB2/firewire enclosures that show corrupt backups. Both use a Prolific PL-3507 chipset. After the latest test following a corrupt firewire backup on my Bytecc ME-340U2F, I decided to remove the drive and connect it directly to the IDE ribbon. The file size was a little over 35GB, so I considered this to be a reasonable test. The data verification in Check Image showed a good image!! I then reinstalled it back into a CompUSA 3.5 enclosure and received the familar corrupt data message.

    Having tried to use both USB and firewire interfaces, it seems the Motherboard's USB and 1394 controller work okay. Data written to the drive appears to be corrupt or not, depending upon whether the connection when attempting to read the image data. The PL-3507 seems to have written accurately when I did the backup as evidenced by the good test on the IDE connection. Also, I have performed other data copy checks using MD5, all of which came back identical to the original hash.

    It seems the Prolific chipset is not allowing the read function to work correctly; or the software check image function is not accurately reading the data stream coming from the PL-3507. Are others having similar results?
     
  15. Mike10538

    Mike10538 Registered Member

    Joined:
    Feb 28, 2005
    Posts:
    1
    RE: I Don't

    I do not have a problem restoring the images with TI-8.0 CWS Build 935/937. I tested it on a DELL GX280 USB 2.0 60G with SATA drive, making the image from windows and restoring with the rescue CD. The results were perfect. I also tested it on a generic PC with 20g IDE USB 1.1, also got perfect results. I backup onto a Segate 160G drive in a basic external USB 2.0 enclosure.
    I am very pleased with ATI.
    Mike
     
  16. brandis

    brandis Registered Member

    Joined:
    Feb 4, 2005
    Posts:
    60
    Did you write and read with Firewire and did get corrupt images and MD5??
    I think you are the first reporting this problem with firewire....

    If reading the data is not o.k. perfect writting is useless.... You can try to find at what file-size this error happen (by zipping some files, copy and do MD5).

    Of cause this is NOT A SOLUTION, only a workaround.
     
  17. brandis

    brandis Registered Member

    Joined:
    Feb 4, 2005
    Posts:
    60
    TO ACRONIS STUFF

    THIS IS A MESSAGE TO THE ACRONIS STUFF!!!!

    Many users with totally different hard/software have this BIG FILE CORRUPTION PROBLEM.

    My oppinion: ACRONIS TRUEIMAGE dont cause the problem!

    But TI is a software which is made to make BIG IMAGE FILES!

    Splitting the image is a only a workaround NOT A SOLUTION!

    WHAT CAN ACRONIS DO ?

    1. ACRONIS should meantion this hardware-problem in the MANUAL and HELPFILES

    2. ACRONIS should add a option in TI for TESTING the quality/accurity of the ports (IDE ATA SATA USB 1394 LAN-DISK-NETWORK SCSI...... whatever we whant to use)
    This test should be made with differnent file-sizes and the results must be written detailed in the LOGS.

    3. the verification must be a option in every (sceduled) image with detailed info written in the LOGS (not only the image is corrupt, but the file XXX.YY in this image is corrupt)

    WHAT DO OTHER USERS THINK ABOUT THIS?)
     
  18. MEDDLER

    MEDDLER Registered Member

    Joined:
    Feb 28, 2005
    Posts:
    3
    Brandis is right. We have heard too little from Acronis in this thread. I, for one, have lost a great deal of critical data because of this fault. Even if, as seems likely, it will be impossible to restore or explore corrupted images, can Acronis suggest a way to retrieve data from these images?
    Come on, Acronis, let's hear from you with some workable suggestions.
     
  19. bocsor

    bocsor Registered Member

    Joined:
    Feb 2, 2005
    Posts:
    8
    Re: TO ACRONIS STUFF

    Brandis,
    I'm still not convinced that the problem is not with TI. Lets consider the case of using the boot cd (Linux) to create and check the image, just to get Windows out of the equation. Keep in mind that there is constantly error checking going on at the BIOS or other firmware level as bytes are written to, for example, a USB hard drive. Cyclical Redundancy Checks are part of the USB data transfer specification for all token and data packets transmitted. Errors would be caught instantly as the data is transmitted. Now you create an image to a USB drive from the Linux based boot cd and you get an "image created successfully" notice when done. You turn around and immediately run the "check image" utility and you can (as I have) instantly get an "image is corrupt" message. I've had this happen every single time on a particular pc, which has never given any problems. I've also gotten this error message so quickly after initiating the check image that TI has obviously not computed a checksum on the entire file. I have then immediately rerun the check image on the same image and this time it progresses several minutes before producing the "corrupt image" message. There is some serious inconsistency in this. That TI should not be able to successfully copy byte for byte from one drive to another I find quite incredible, especially since no errors are generated during the process. There are clearly some machines that just will not produce an image that is not "corrupt" no matter what one does. However, I still think the fault is in the way that TI either checks the image after it is created, or in the way it is created in the first place - on these machines. The real question is "what is the significant difference in these machines?", and I don't think its faulty hardware.

    Regards,
    George
     
  20. wdormann

    wdormann Registered Member

    Joined:
    Jun 27, 2004
    Posts:
    480
    boscor:
    You must be 100% sure to eliminate any possibility of hardware / OS / driver issues before even attempting to look into issues with a particular application.

    Have you tried a straight file copy of a multi-GB file in Windows to the device you're backing up to? Does the MD5 match every single time?
     
  21. bocsor

    bocsor Registered Member

    Joined:
    Feb 2, 2005
    Posts:
    8
    Actually, the "corrupt image" issue I was referring to happens on this particular pc even when the image is written to a folder on the one and only internal IDE drive on the machine. There is no transfer of the image file(s) going on here. It is written to the internal hard drive directly and immediately checked. Of course, in this case it was written from within the Windows version of TI. I've tried every conceivable senario on this machine- internal, external USB- it just will not work. On other machines TI works just fine. If there is truly a corrupt image, then TI should in no way have generated the message "your image was successfully created" in the first place, as it does every time.
     
  22. wdormann

    wdormann Registered Member

    Joined:
    Jun 27, 2004
    Posts:
    480
    Have you run memtest86 overnight without errors?
     
  23. bocsor

    bocsor Registered Member

    Joined:
    Feb 2, 2005
    Posts:
    8
    I've run memtest86 for longer periods and more often than TI (but not overnight). Also ran thorough scandisk checks numerous times - no problems. Apparently TI can find problems if run just once ?? Just kidding, sort of. The hardware is fine, or at least good enough to just by chance run a successful TI check image once out of the dozens of times I've tried it. Basically, I agree with brandis' last post that TI needs to be altered with respect to this problem.
     
  24. Menorcaman

    Menorcaman Retired Moderator

    Joined:
    Aug 19, 2004
    Posts:
    4,661
    Location:
    Menorca (Balearic Islands) Spain
    Wholeheartedly agree. I too have suggest that users experiencing image corruption problems should carry out a Windows multi-Gigabyte file transfer across USB, check the MD5 checksum of the original against that of the copy and post back their results. To date, very few people have done so. Is it because they don't possess a suitable MD5/hash utility? If so, here is a link to <Easy MD5 Creator>.

    Regards
     
  25. Ed Y

    Ed Y Registered Member

    Joined:
    Oct 18, 2004
    Posts:
    86
    bocsor,

    I can only talk to my situation but when the image is created by TI (about 5 gb's) to spare space on my C: (primary) drive, then verified by TI Check Image, it will verify everytime. If I then compute the MD5, use Windows to copy it over to a TEAC USB 2.0 external hard drive (uses WD internal), 90% of the time the MD5 will be different and the TI Check Image run will also show corrupt, usually towards the ed of the run.

    Can't see how this type of error has anything to do with TI.

    Now if I let TI create the image on the external USB drive (about 5 gb's), then once again Check Image will show corrupt about 90% of the time.

    If I create the image on the external USB drive using 700MB as the image file size, a Check Image run will always verify OK or at least it has so far. I've run the memory tests, both single and overnight and nothing showed.
     
Thread Status:
Not open for further replies.
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.