Acronis TrueImage 11 Home : Corrupt Images!

Hi all,

I used to be a semi-frequent participant in this forum. e.g. I put together instructions for how to make a bootable rescue DVD for ATI.

I recently saw a deal on ATI Home 11, so I decided to purchase the upgrade. I've been using ATI 8.0, but figured it'd be nice to have updated hardware support and some of the new backup features.

I tried a simple backup of my C: drive and tried validating the image using the rescue CD. It complained about a corrupt image. So I tried validating from within Windows. Corrupt image. I tried creating an image to my second internal IDE hard drive (instead of external SATA drive). Validation indicates that the image is corrupt. I tried a different split size (4.7GB preset instead of manual 1492MB). Corrupt image.

The logs show these sort of errors:
Pending operation 3 started: "Verifying backup archive"
The archive is corrupted: None
Operation with partition "0-0" was terminated.
Details:
Image corrupted (0x70020) Tag = 0xF5F8CBCF76155638>

I have not tried actually restoring the image for obvious reasons. I haven't followed the status of TrueImage for years, but has it really actually gotten worse than previous versions? (Like I said, I've been using 8.0 with minimal trouble)

Do I need to start the refund process?

 

Just some more evidence regarding this problem:

I took a known-good image that was taken with ATI 8, and this image resides on a network (CIFS) share. I tried the following tests:
Verify with ATI 11 from Windows: Corrupt
Verify from ATI 8 Rescue CD: OK
Verify from ATI 11 Rescue CD: Corrupt

I find it hard to believe that ATI 11 is just flat-out broken, but all the evidence seems to be pointing to this!

 

You've pretty well ruled out a problem with the archive storage device. It is possible that TI11 is using more/different RAM than TI8 and there is a bad location. I'm running TI11 for basic imaging with no problem at all.

IMO, most of TI11's problems are associated with the "clever features", not the basic imaging engine. Naturally, the Linux rescue CD driver issues still get some users but you have problems in Windows too.

I'd download free Memtest86+, V2.01, from www.memtest.org (note the .org) and see if it sheds any light on the situation. Ideally, it should run overnight.

 

Hi seekforever,

Thanks for the response. I'm starting to feel a bit silly here. I have run memtest86+ before in the past, and I have a system with ECC RAM with the ECC option enabled in the BIOS.

However, I just re-ran the test and I saw two things of interest:
1) Memtest86+ indicates that ECC is not enabled. (A8N-E BIOS bug?)
2) The memory showed errors around the 900MB mark (out of 1GB)

Sooo, without reliable hardware I obviously can't blame ATI 11 in this case. The ATI 8 working vs. ATI 11 not working sure is a red herring, though!

Thanks!
-WD

 

As seekforever suggested, if the farther reaches of memory are not employed, they won't produce errors. It's likely that ATI8 doesn't use as much memory space as the newer versions -- 8 is pretty old.

 

Glad to be of help. On the above ECC issue, did you run the 2.01 version or perhaps an older version. If older, it could be that it doesn't understand your chipset and is not reporting correctly. I would guess if it displays the correct chipset number in the Memtest86+ screen it is supported.

 

OK, I've got some not-so-reassuring details here:

I took out my second memory stick, and re-ran the memtest86+ 2.01 test for about 8 hours without error.
I tried verifying the images with ATI 11 (in Windows and from rescue CD, verifying from various sources) and they came up with the same error message saying that they were corrupt. It's important to note, though, that when an image verifies as corrupt, the error dialog comes up at the exact same point in the verify process. i.e., it's in no way random or based on other things that the computer is doing.
Verifying the old images with the ATI 8 rescue CD shows that they are fine. I can't verify the newly-created ATI 11 images with the ATI8 CD, due to backward compatibility issues.

I put the second memory stick back in (and shifted both over one slot for the hell of it) and it passed memtest86+. Prime95 tests within Windows show that everything is copacetic. Same verify results with ATI 11, though.

So I'm coming up with two possible theories, neither of which are making me feel very comfortable at all:

1) The "known-good" (i.e. successfully validated and restored from) images that I created with ATI 8 are actually not good (due to flaky hardware or whatnot), and the reason that ATI 11 is showing them as corrupt is because Acronis has improved the image validation feature between versions 8 and 11.
2) ATI 11 really is just kinda broken and I should go back to ATI 8 and be done with it.

What's really concerning to me is that both theories involve ATI creating images that fail to validate, but no indication of any error is given at the time of image creation.

Any thoughts?

 

Life isn't simple,alright!

If we assume that maybe the memory did cause a problem (although it isn't certain) it could be that all the TI11 images you made are bad. Maybe you did this but since the situation isn't clear, I'd pull the "bad" stick again, make an image with TI11 and see if it validates.

In case you aren't aware, TI's method of validation is to write a checksum for every 256K bytes of data when the archive is written. The validate procedure reads the archive and recreates the checksums from the read data in memory, if even one bit out of 4000 checksums/GB is bad the archive is declared corrupt.

You might say that if the checksum is based on bad data in memory and the bad data and its corresponding checksum were written to disk, read properly, it would not give an error since the data read would be the data the checksum is based on.

I am not aware of TI11 not validating TI8 archives because of format changes. Acronis says they guarantee one version backwards compatibility but their record has always been better than that. There was a format change from V8 to part way through V9 and then in V11.

 

I'm running some further tests, so it'll take some time to come to a conclusion, but here's where I'm at so far.

- I had a system with an A8N-E motherboard, and I've been successfully using Acronis TrueImage 8.0. Backups validate and restore fine.
- At some point, this motherboard failed, and I purchased the same model off of eBay used, as it was discontinued and didn't want to switch brands/models.
- I continued using ATI 8.0 and everything seemed fine.
- Now since updating to ATI 11, I noticed the image validation problem and also an intermittent memory problem. Several test images were created with ATI 11 with possibly-bad RAM.

My last test was to create a new image with my re-socketed RAM modules. ATI 11 was able to create and validate the images without error. (I've actually done 4 sets of create/validate pairs: one from the rescue CD and 3 from within Windows). And I took some of the images that were created before I replaced my motherboard, and those too validated fine. However, first couple of images that I created with ATI 11 still come up as corrupt.

So my current conclusion is this: (isn't it fun how these change!)
Images that were created with my "new" motherboard before moving my RAM sockets have a good chance of being corrupt. I've had some slight instabilities with this board, but nothing that I could reproduce, and very intermittent (so nearly impossible to troubleshoot). So creating corrupt images seems plausible. That plus the fact that the corrupt image error message comes up in the exact same spot for each image set. If it were a memory flaw during verification, I'd think it would be more random. It's very disturbing that ATI 8 didn't/doesn't detect this problem, but it's the best that I can come up with at this point. The common thread with images that fail to validate with ATI 11 is that they were written using ATI 11 or ATI 8.0 with my new motherboard, before I mucked with the RAM.

I'm not exactly comfortable with a system that is "fixed" by moving RAM slots, so in the long run I'll probably replace the board and RAM. But for now I think that setting up ATI to automatically validate images after creating them will suffice.

 

It seems you are on the right track although you never know for certain.

Do you think that your instabilities could have been the result of a RAM problem? Possibly an intermittent connection in the RAM connector? Of course, it is possible there is a problem with a connection (solder joint, plated-through hole, wire-bond inside a chip, ...) on a particular stick. That's the problem when things get fixed by moving, you never know exactly what mechanism has changed something.

Turning on the Validate is a good idea. I always say the TI validation is pretty good system test! I would remove the RAM stick you suspected when it happens and try again to see if it pins anything down.

 

Yeah, that's my best guess at the moment. Since switching boards, I've had a few BSODs. Looking at the crash dump files, the cause seemed to be different each time. I updated the seemingly-related drivers and improved airflow, but the crashes would be infrequent enough to know if I fixed the problem.

I guess ATI 11 is acting like a canary for me! Only time will tell if it's a permanent fix...

Thanks for all the help.

 

Intermittent RAM problems can be very elusive. I had a problem with a module in my Linux PC that would produce a spurious lockup maybe once every 2-3 months on a machine running 24/7. It took a week of running Memtest86+ to produce even a single error! After cleaning the contacts of the module and reinserting it, the machine ran fine for 3 more months so I thought it was fixed. Then it happened again. When RAM prices dropped to $15 I just replaced the module. That was 6 months ago so I think that was it.

 

You can have mem probs that only show up when the mem reaches a certain temp, this can be due to flaws inthe solid state devices and it can be the elctrical contacts where the mem module meets the mobo. The contacts are also affected by temperature and a just good enough connection can become just barely bad at the right temperature.

 

Hey...here's a curve ball that I've found when using the Seagate version of both ATI10 and 11. On only one machine (a Shuttle machine that has worked ok as an anti-virus server in our Win2003 domain) I have the same error message come up during the verification step...and here's the important part...ONLY when the target drive is connected via USB. I've tried at least three different USB-to-IDE connectors and at least two different drives and always get the same error. However, if I backup to an otherwise unused partition on the system drive, validation passes with flying colors.

There are many things that I have NOT tried on this rig, such as imaging the entire drive to another drive on the internal drive controllers, using a for-pay version of ATI, eliminating the possibility that the validation error comes from backing up that "otherwise unused partition", etc.

Since I can survive by imagining to an unused partition and offloading the resulting images after the b/u, I just haven't taken the time to explore things more deeply. I had hoped that the latest/greatest versioln from Seagate (v11.0 build 8142) would address some nasty USB issue that the Shuttle board presented, but no dice.

FWIW: Although we use the for-pay server version of ATI on the mission-critical production servers (wouldn't be without it...I sleep much better at night knowing I've images in case of fire) I've found that the freebie Seagate version has worked in every instance except this one machine. It makes a compelling reason to stick with Seagate/Maxtor drives for us on our desktops/laptops.

So...what happens if you target a drive externally? (Just the reverse of my situation.) I'm assuming a lot here, I know....

- Joel