Error 70020 Image corrupted- hardware?

Hi all,

I am hoping someone can help. I recently started using Acronis Home Edition 9.0. First couple of backups worked fine - took an image of my entire C drive (system drive only - I have a seperate data drive) - that worked fine. A few days later I needed to use it to restore the drive - and it worked like a dream.

Here is my problem - sometime shortly after that I started getting Validation errors (70020) whenever I tried to create a new backup. I have performed chkdsk /r on all my drives with no problems. I have run memtest overnight with no problems. However I still get validation problems. I get validation problems even when I make the backup when booted from the rescue disk and I see checksum problems if I simply copy large (~4GB) files in Windows (ie. MD% checksums of original and copied files do not match). From what I can gather from these forums this implies a hardware problem, but I am unsure where to look next. Any suggestions?

Regards,

Sean

 

Look in the XP Event Logger's System Log and see if there are any suspicious entries in it.

I had problems with SATA drive cables that gave validate errors but seemed OK for regular PC operations. There were entries in the System Log; I believe Parity Errors.

If that doesn't help, I would still consider memory a possibility even if Memtest ran overnight (running the diagnostic is not exactly the same as running the system). Do you have more than one stick such that you can try them one at a time?

Are your memory timings set aggressively or the standard default values? Are you overclocking the system in any way?

Any electronics person will tell you to always check the voltages first; so that is never a bad idea.

Do your checksum errors happen when copying between the two drives, or the copying to the same drive?

 

Hi Seekforever,

Thanks for replying... I will definitely check the event log - I have had a quick look previously and not found anything - but will look again.

Regarding oc'ing and voltages - I have not touched anything - I have a DFI Infinity RS-482 that was running with 2x256 DDR266 and AMD 3500+ for a few months without any problems. I have recently upgraded to AMD 4600+ X2 with 2x512 G.Skill DDR400 and things have not been quite right since. The voltages all look ok - but I will try dropping my old ram back in and see if that makes a difference.

Checksum errors seem to happen even if copying between two directories on the same drive.

Regards,

 

Greetings Merlin1953,

Your checksum errors in Windows confirm that, as on my system, the errors are not TI errors, but hardware/driver issues. I did a bit of research on your DFI Infinity RS-482 motherboard (check the reviews on Newegg, for one) and several have reported having problems with a bad memory slot, and with the SATA and USB chipsets. I had lots of problems early on with what I thought was a mainstream, albeit older, Gigabyte mobo, and it turned out my chipsets weren't all that robust, either. So, you might consider your mobo...not good news, I know, but, finding the cause is always best. You could also test your memory one slot at a time, to see if that narrows down the options.

TI definately puts chipsets and subsystems thru a stress test -- copying multiple gigabyte images at high speeds reveals flaws that everyday computing doesn't. I found using different drivers helped my system run better...you might experiment with those...on my system, the mobo nvidia drivers for SATA caused problems that don't occur using Microsoft drivers.

A few reviewers also reported problems with the DFI Infinity RS-482 mobo running under Linux. If TI can't run your system in Linux mode, you could try the Bart PE option, discussed in other threads in these forums. Though some TI operations, such as restoring a system drive, rely on the Linux mode rather than Windows, so, be sure to test each step of the Archive creation/restoration process on your system, to be certain it works.

I was born in '53 also, if that's the meaning of your 1953...a good year!

Regards

 

Hi Christopher NC,

Thanks for that... Am starting to suspect more and more that you are right - that it is the mb - it would be nice to find a definitive test for it rather than just by process of elimination.

Just for information (in case any others end up trawling through this forum with similar issues....). The mb with 3500+ CPU and 2x256MB ram ran Ubuntu (Dapper) quite well - with the exception of the sound which I could never get working.

you are right - the only reason this is coming up in a TI forum is that it does seem that TI causes some hardware stress that would otherwise go unnoticed... Although my wife has just started have problems editing large movie files.....

'53 is the rego number of my boat - not my year of birth

 

Hi all,

Again just in the interests of complete disclosure....

I tried dropping in completely different RAM sticks into completely different slots, and tried resetting all BIOS settings (both fail safe and optimised) and none of it helped. (i.e. a checksum compare of a large (~4GB file) after copying between directories on a single physical disk still fails). This leads me to Christophers conclusion that it is my mobo

Regards,

Sean

 

I'm inclined to agree. Sounds like it's similar problem an old VIA chipset model had years ago - bad large-file transfers. The fact your wife is having problems with video probably is the same thing since they tend to be large files also.

Did it start exactly after your CPU/mem upgrade? If so, could be that pushed something over the edge and may be fixable by specifying a lower clock frequency (which I'm sure is not your idea of a great thing).

Any BIOS updates?

 

Seekforever:

This is a bit off thread, but, since you understand these things, would chipset errors show up consistently, as in when copying large files in Windows and in all modes of TI? Or, could these errors be driver dependent, and only show up while, for example, the system is being run by TI Linux Boot mode?

I ask, because, if my hardware is to blame, I'd like to know, since I just posted again about the Boot Mode files and folders corruption I've been experiencing, and which seems to happen only to some, perhaps even a few, users? Could faulty chipsets, or drivers, be the culprit? Or, does TI just not know how to drive cars with finicky transmissions, with its Linux boots on?

Do you know of a reliable testing program that can ferret out such errors?

 

The problem is the answer can be the driver and it can be the chipset design or a hardware failure.

Obviously the Linux drivers are not the same as the Windows drivers and it is possible for something to work in one and not the other. It may not be because one driver is written wrong but that it is using a particular feature of the hardware that is problematic. Also, if something is on the edge it could be that one driver due to the way it is written has a bit more of a pause between critical instructions which gives the HW more time to get into the right state. I have seen (non-PC, but wouldn't be surprised it they were in them as well) drivers with NOPs (No Operation) instructions inserted into them for this purpose. Not great design necessarily.

If both OSs fail then it does put emphasis on something being wrong with the HW. I am reluctant to say any particular part of the HW because it really all has to work together as a system.

When doing checksum calculations pretty well all of the major components of the system are being used, CPU, memory sticks, memory bus, disk drive, disk controller, PCI bus, etc.

I am not up on the available diagnostic programs to make any recommendation.

I would investigate the MB manufacturer's website and any forums for that MB looking for reported problems that are similar. You could also make use of their support services to ask the question. You likely can find a newsgroup for the MB on the Usenet newsgroups.

 

Hi Seekforever,

I was doing some more testing last night and noticed something that seemed to be TI specific...

When I copy video files, I seem to be able to copy files that are many hundreds of MB in size before I start seeing checksum problems. However when I copy TI image files, it seems as though I get errors on files that are only a few MBs in size.

Do you know of any TI specific reason for this?

Thx and regards,

Sean

 

If you are referring to copying both the video and TI files from one location to another using Windows Explorer and running the checksum calculation before and after copying then I don't see any reason for it other than perhaps a bad area on the disk. However, you have run chkdsk /r on the partitions.

If you have a normally compressed TI file I think it would look very similar to a video file which is usually compressed. I really can't think of anything.

 

Thanks seekforever,

That is exactly what I am referring to - and like you I thought that copying a large tib file and copying a large video file would be similar. (It does at least provide a glimpse of hope for me to continue bashing at this. I would like not to have to upgrade my mb - and if I can get ATI working on ~600MB files (which is what a Windows copy seems to be able to handle) then I may at least have a feasible workaround jsut by splitting the image files.)

Regards,

Sean

 

I don't mean to dash your hopes, but early on, while I was trying to get TI to work on an Ext USB HD, I tried this method. I split files, and tried smaller and smaller files, to no avail. My problem was hardware based, and solved with an Adaptec USB PCI card, and a different USB enclosure. You may or may not have similar results. But, if, like me, you don't want to change your motherboard yet, and can afford one or two Adaptec controller cards (even Comp USA carries them), I found that they worked like a charm on my system.

Once upon a time (my 2nd Windows 98 system still has Ultra Wide SCSI) it was common to use PCI controller cards to run Hard Drives. So, I thought I'd give it a try, by-passing at least some of the SATA & USB chipset issues on my motherboard. NEC chipsets are reputed to be the best for controller cards, by the way...and I'm very happy with Adaptec's build quality.

My other speculation, and it's purely that, is that TI seems to do things at maximum speed, while other programs don't always. With DVD burns, TI often fails to verify, but several have reported luck copying the same DVDs at slower speeds, and then having the Archives verify. If, while copying large Images files to and from hard drives TI also uses the top speed your hardware reports as available, that might explain why TI has errors that Windows doesn't using the same hardware.