This error message often occurs a short while after an image is launched. The context is the following.

It happened in two phases. One is over the last days, and the other one was 2 years ago. I cannot vouch the error message was the same the last time, but the rest of the context was. I don't know whether it's relevant, but it might.

Each time, imaging attempts fail repeatedly. After a while, normal operation resumes. Right now, my last attempt was unsuccessful.

Each time, it happens shortly after (or before) one of my external backup disks, linked by USB 3, becomes RAW and unreadable.

Each time, this is accompanied at some point by blue screens.

Right now, I have regained access to my external backup disk, through CHKDSK /F. All images were there. Fulls and differentials could be mounted and browsed, hower incrementals could not. A fresh full image was made. It could be mounted, and browsed. However, Macrium said the backup failed, because it could not be verified.

At that point, I decided to do a full format of the drive. Then I attempted a full image again. That's when I got the above error message, which I also got several times during the last few days, while I was trying to solve this RAW problem, when trying to backup on my other external disk. The last imaging attempt was done through an eSATA cable.

I believe the formerly RAW disk is physically sound. The critical SMART attributes are good, and it passes the short SMART self-test.

Thank you for your help.

Windows Home Premium 64-bit, homemade computer.

Try CHKDSK /R rather than /F. That will cause CHKDSK to scan the disk surface looking for bad sectors. It will take quite a bit longer than a /F operation though, so plan accordingly. I believe a full format also includes a surface scan, but practically nobody does those anymore because they take forever on large drives, which is why the Quick Format option is checked by default (but doesn’t include this scan).

jphughan - 5 June 2020 1:46 AM
Try CHKDSK /R rather than /F.

Why ? I've already done a full format. Does this error message indicate bad sectors on the target disk ?

Anyway, since I published my post, I made two new fresh attempts at imaging, which produced the same error message early in the process. One on that formerly RAW disk, by USB 2 cable, and one on my other, good backup disk, by USB 3 cable.

It's quite unlikely that both disks developed bad sectors to the point of failing Macrium. Also, the once-RAW disk has good SMART scores, and a 100 % bill of health by Hard Disk Sentinel.

Does anybody know what that error message means ? I searched the forum for it and found nothing.

+x

 

Clairvaux - 5 June 2020 9:05 AM

jphughan - 5 June 2020 1:46 AM
Try CHKDSK /R rather than /F.

Why ? I've already done a full format. Does this error message indicate bad sectors on the target disk ?

Anyway, since I published my post, I made two new fresh attempts at imaging, which produced the same error message early in the process. One on that formerly RAW disk, by USB 2 cable, and one on my other, good backup disk, by USB 3 cable.

It's quite unlikely that both disks developed bad sectors to the point of failing Macrium. Also, the once-RAW disk has good SMART scores, and a 100 % bill of health by Hard Disk Sentinel.

Does anybody know what that error message means ? I searched the forum for it and found nothing.

Thanks for posting.



The error is generated by the Windows IO sub system and is the result of a bad sector (Cyclic Redundancy Check) reported by the source disk of the image when the disk data cannot be read reliably. More info here:

https://knowledgebase.macrium.com/display/KNOW72/Imaging+disks+with+bad+sectors
https://docs.microsoft.com/en-us/windows/win32/debug/system-error-codes--0-499-
https://devblogs.microsoft.com/oldnewthing/20130711-00/?p=3833

Kind Regards

Nick

Macrium Support

See our reviews on

Thank you, Nick. Very interesting links.

Meanwhile, I had suspected as much, ran CHKDSK /R on the C: partition of one of my two internal drives, and the subsequent image completed successfully. A few questions, though. The KB says :

A disk with bad sectors is permanently damaged.
If a bad sector is encountered during image creation then you receive this error in your backup log:
Backup aborted! Unable to read from disk - Error Code 23 - Data error (cyclic redundancy check).

• Does bad sector here exclusively mean physically bad sector, as opposed to logical ? And is this error message exclusively generated by physical bad sectors ?
• I reckon activating the option to ignore bad sectors, when creating an image, makes the image less reliable, should be used as last resort only, and would be especially risky for system partitions. Is that correct ?
• Can bad sectors on the imaged disk somehow "export" to the backup disk, and make it turn RAW ? In the two instances the latter happened to me, it more or less coincided with bad sectors on the imaged disks, and blue screens. Or is there no relationship ?

+x

 

Clairvaux - 6 June 2020 9:03 PM

Thank you, Nick. Very interesting links.

Meanwhile, I had suspected as much, ran CHKDSK /R on the C: partition of one of my two internal drives, and the subsequent image completed successfully. A few questions, though. The KB says :

A disk with bad sectors is permanently damaged.
If a bad sector is encountered during image creation then you receive this error in your backup log:
Backup aborted! Unable to read from disk - Error Code 23 - Data error (cyclic redundancy check).

• Does bad sector here exclusively mean physically bad sector, as opposed to logical ? And is this error message exclusively generated by physical bad sectors ?
• I reckon activating the option to ignore bad sectors, when creating an image, makes the image less reliable, should be used as last resort only, and would be especially risky for system partitions. Is that correct ?
• Can bad sectors on the imaged disk somehow "export" to the backup disk, and make it turn RAW ? In the two instances the latter happened to me, it more or less coincided with bad sectors on the imaged disks, and blue screens. Or is there no relationship ?

Thanks for getting back.


It's irrelevant whether a physical or logical sector read fails. The disk is unable to read the requested block.

All logical sectors map to physical sectors. Physical  sectors map to the underlying storage media which may or may not be contiguous physical storage. It really doesn't matter. The Raymond Chen blog linked above does a good job of explaining how CRC errors 'bubble up' from the disk.


A disk image is an exact representation of the file system at the time the image is taken. Bad sectors cannot be read so the image has the same 'reliability' as the file system being imaged.

If you have bad sectors then it's time to replace the disk.


No, this can't happen. There's a problem with the backup target disk or the data transport to the disk.

Kind Regards

Nick

Macrium Support

See our reviews on

"If you have bad sectors then it's time to replace the disk."

Thanks for the reply. Sorry I'm late, I did not receive an email alert, don't know why. Apologies for not finding the way to quote properly.

I have one "damaged sector" on my system disk (says Hard Disk Sentinel), but this comes and goes. Sometimes the disk auto-corrects it, and it disappears, sometimes it reappears.

I applied the Repair feature of Hard Disk Sentinel, and this brought back the disk to a pristine state (according to the same program). Now this "damaged sector" (not "bad", according to the software's lingo), has cropped up again.

Should really a disk be discarded because it has one single bad (or otherwise defective) sector ? I was under the impression that a certain amount of bad sectors was considered normal, and not the end of the world per se.

Something out of the ordinary appears to be going wrong.  When a sector is marked as "bad," it should be completely ignored for all future write requests.  Why a sector would be marked as bad and then require it be marked bad again, I couldn't say.  To be honest, I wouldn't necessarily say the drive is going bad.  Something may be marking and unmarking that sector as bad repeatedly.

+x

 

TimotnyM - 19 July 2020 6:01 PM

I have another observation regarding this phantom read error. If indeed it is Windows throwing the error, one would expect Windows Event Logs to record it. On my system, there is no such disk read error logged. Since I can't find any disk issues, not even remapped sectors/block within my disk allocation tables, or reported as such by my SSD tools, and Windows says there haven't been any disk read errors, I must assume the problem is Macrium Reflect Home 7.2 ... and NOT a real disk issue.

Thanks for posting.



This is a Windows error and is in response to the ReadFile Windows API when reading the disk directly or from the volume shadow copy. It's 'bubbled up' from the disk I/O sub-system and ultimately originates from the disk firmware. It is very strange if there are no events logged but you really don't want to ignore it or assume that it isn't real.

Raymond Chen has an excellent blog on this error here:

https://devblogs.microsoft.com/oldnewthing/20130711-00/?p=3833

Have you run chkdsk /r on the volume with the error? The '/r' switch is important as it scans the entire volume for bad sectors. 

Kind Regards

Nick

Macrium Support

See our reviews on

This was an SSD showing up with a CRC error, correct?  I forget.  If so, what make and model?  I ask because if it's a specific one, there was a post somewhere else about an SSD that was returning weird Reflect errors until a disk utility included by the manufacturer of the SSD was uninstalled.  Reflect performed fine on it with that utility uninstalled.