Question Question about random bad sector warning from Macrium Reflect backup that nothing else can find.

[Darkmatterx]

When I do a full image of my system drive I have a about a 40% chance of getting a warning that I have one bad sector or more. Since it's only some of the time I'm assuming that that space is sometimes not occupied with data, or Macrium is wrong. I suggest possibly wrong because no other method of checking the M.2 SSD shows a problem.

Oddly enough, the problem got better, or detected less, after the first few logs, despite programs like chkdsk and even Samsungs own SSD "health" software never finding a problem. I also tried some other 3rd party programs which also found nothing.


The model is a Samsung 970 EVO 1TB.

Other hardware includes:

AMD Ryzen 7 5800X
GIGABYTE X570 AORUS PRO WIFI
EVGA RTX 3080 Ti FTW3 ULTRA GAMING
2X 32GB Crucial (DDR4-3603 / PC4-28800 DDR4 SDRAM UDIMM)

OS: Win10 Pro 64bit


Since there's no support for the free version, I was hoping someone here might be able to offer some assistance.

Each space represents a different full backup image. Newest at the top, oldest at the bottom; Sorry about that.

The very bottom one freaked the hell out of me, but when I did another full backup the very next day it "only" found 4 clusters. The problem seems random except for cluster 47841470 and a few other clusters that may show up a few times. Many clusters never show up again.

I also never get warnings from my Differential backups. Lastly, these are intelligent full backups. I may try doing full backups that also copy unallocated space to see what sort of errors I may get. All backups are verified at the end.

In total, I currently have logs for 33 full backups. 13 of which (all included in that count, including ones omitted below) had a warning of 1 or more errors. That's barely under 40%.

Why can no other program detect this? I fully admit, that I only just now realized that I didn't try a different free full image backup program, but I figured at least Samsungs own software should detect something! BTW, since it will be asked, SMART data is all in the green.

Thanks!!!


The logs...
Skip to the bottom if you want to read these in date from oldest to newest. My bad. I also omitted a couple of the ones that only showed cluster 47841470. I'll sometimes go a month without seeing a single error, despite doing a full backup weekly. I'll just mention one more time that all of these are intelligent full backups.

2025/04/23 (In this, the most recent full backup, and scattered throughout my logs, is the cluster I see the most, 47841470. Some omitted from this list for length.)
Read Error: Bad sector found in cluster 47841470 (Most often seen cluster)


2025/03/17
Read Error: Bad sector found in cluster 47841470
Read Error: Bad sector found in cluster 47846078

2025/01/31
Read Error: Bad sector found in cluster 81578353

2025/01/24
Read Error: Bad sector found in cluster 47841470

2025/01/17
Read Error: Bad sector found in cluster 74492545
Read Error: Bad sector found in cluster 130317216
Read Error: Bad sector found in cluster 130317237
Read Error: Bad sector found in cluster 130317257

2025/01/10
Read Error: Bad sector found in cluster 47841470

2024/12/06
Read Error: Bad sector found in cluster 157951385

If you read this from the bottom up, it was at this point that I was both very confused and very concerned. Someone else online suggested that it's possibly a mistake or problem on Macrium's end since the number of bad sectors varies so wildly. In fact, as I said near the top, I only get an error message about 1/3 of the time. The lack of errors found in any full image between Oct 12th and Dec 5th was also odd. I have gaps like that throughout my logs.


2024/10/11
Read Error: Bad sector found in cluster 34315131
Read Error: Bad sector found in cluster 47841470

2024/09/09
Read Error: Bad sector found in cluster 34315131
Read Error: Bad sector found in cluster 47841470
Read Error: Bad sector found in cluster 72455511
Read Error: Bad sector found in cluster 72455523
Read Error: Bad sector found in cluster 72455526
Read Error: Bad sector found in cluster 72455528
Read Error: Bad sector found in cluster 72455529
Read Error: Bad sector found in cluster 73875783

2024/08/28
Read Error: Bad sector found in cluster 61732897
Read Error: Bad sector found in cluster 73875783
Read Error: Bad sector found in cluster 95141323
Read Error: Bad sector found in cluster 151633627


2024/08/27
Read Error: Bad sector found in cluster 47841470
Read Error: Bad sector found in cluster 58701130
Read Error: Bad sector found in cluster 58701131
Read Error: Bad sector found in cluster 58701135
Read Error: Bad sector found in cluster 58701142
Read Error: Bad sector found in cluster 58701188
Read Error: Bad sector found in cluster 58701193
Read Error: Bad sector found in cluster 58701197
Read Error: Bad sector found in cluster 58702302
Read Error: Bad sector found in cluster 58838379
Read Error: Bad sector found in cluster 58838388
Read Error: Bad sector found in cluster 63369331
Read Error: Bad sector found in cluster 63369338
Read Error: Bad sector found in cluster 64044479
Read Error: Bad sector found in cluster 67484519
Read Error: Bad sector found in cluster 67886827
Read Error: Bad sector found in cluster 70704477
Read Error: Bad sector found in cluster 70705661
Read Error: Bad sector found in cluster 70713489
Read Error: Bad sector found in cluster 70713501
Read Error: Bad sector found in cluster 70722793
Read Error: Bad sector found in cluster 71782679
Read Error: Bad sector found in cluster 71783627
Read Error: Bad sector found in cluster 71783656
Read Error: Bad sector found in cluster 71785779
Read Error: Bad sector found in cluster 71785895
Read Error: Bad sector found in cluster 71785923
Read Error: Bad sector found in cluster 71785942
Read Error: Bad sector found in cluster 72455511
Read Error: Bad sector found in cluster 72455523
Read Error: Bad sector found in cluster 72455526
Read Error: Bad sector found in cluster 72455528
Read Error: Bad sector found in cluster 72455529
Read Error: Bad sector found in cluster 72455963
Read Error: Bad sector found in cluster 72455992
Read Error: Bad sector found in cluster 72456283
Read Error: Bad sector found in cluster 72456287
Read Error: Bad sector found in cluster 72456290
Read Error: Bad sector found in cluster 72456292
Read Error: Bad sector found in cluster 72456293
Read Error: Bad sector found in cluster 72457911
Read Error: Bad sector found in cluster 73875783
Read Error: Bad sector found in cluster 79704544
Read Error: Bad sector found in cluster 79704566
Read Error: Bad sector found in cluster 79704587
Read Error: Bad sector found in cluster 79705220
Read Error: Bad sector found in cluster 79705224
Read Error: Bad sector found in cluster 82314358
Read Error: Bad sector found in cluster 82321270
Read Error: Bad sector found in cluster 83075946
Read Error: Bad sector found in cluster 83109622
Read Error: Bad sector found in cluster 84949419
Read Error: Bad sector found in cluster 84949423
Read Error: Bad sector found in cluster 84949426
Read Error: Bad sector found in cluster 84949428
Read Error: Bad sector found in cluster 84949429
Read Error: Bad sector found in cluster 84949437
Read Error: Bad sector found in cluster 84949444
Read Error: Bad sector found in cluster 84949450
Read Error: Bad sector found in cluster 84949455
Read Error: Bad sector found in cluster 84949459
Read Error: Bad sector found in cluster 84949462
Read Error: Bad sector found in cluster 84949464
Read Error: Bad sector found in cluster 84949465
Read Error: Bad sector found in cluster 84956043
Read Error: Bad sector found in cluster 84962283
Read Error: Bad sector found in cluster 84962313
Read Error: Bad sector found in cluster 84968547
Read Error: Bad sector found in cluster 86720752
Read Error: Bad sector found in cluster 86720786
Read Error: Bad sector found in cluster 86720880
Read Error: Bad sector found in cluster 86720904
Read Error: Bad sector found in cluster 86721776
Read Error: Bad sector found in cluster 86721803
Read Error: Bad sector found in cluster 95952115
Read Error: Bad sector found in cluster 95952140
Read Error: Bad sector found in cluster 95952199
Read Error: Bad sector found in cluster 95952887
Read Error: Bad sector found in cluster 95953651
Read Error: Bad sector found in cluster 96618230
Read Error: Bad sector found in cluster 96618244
Read Error: Bad sector found in cluster 96618257
Read Error: Bad sector found in cluster 96618269
Read Error: Bad sector found in cluster 96618280
Read Error: Bad sector found in cluster 96618822
Read Error: Bad sector found in cluster 96619402
Read Error: Bad sector found in cluster 96619415
Read Error: Bad sector found in cluster 98632144
Read Error: Bad sector found in cluster 98632169
Read Error: Bad sector found in cluster 98838024
Read Error: Bad sector found in cluster 98838037
Read Error: Bad sector found in cluster 101763294
Read Error: Bad sector found in cluster 101763306
Read Error: Bad sector found in cluster 101763418
Read Error: Bad sector found in cluster 101763428
Read Error: Bad sector found in cluster 106141694
Read Error: Bad sector found in cluster 106141723
Read Error: Bad sector found in cluster 106142290
Read Error: Bad sector found in cluster 106142294
Read Error: Bad sector found in cluster 106142297
Read Error: Bad sector found in cluster 106142299
Read Error: Bad sector found in cluster 106142320
Read Error: Bad sector found in cluster 106142340
Read Error: Bad sector found in cluster 109128017
Read Error: Bad sector found in cluster 109128085
Read Error: Bad sector found in cluster 109128104
Read Error: Bad sector found in cluster 109128114
Read Error: Bad sector found in cluster 109128205
Read Error: Bad sector found in cluster 109128218
Read Error: Bad sector found in cluster 110135009
Read Error: Bad sector found in cluster 112287737
Read Error: Bad sector found in cluster 112577486
Read Error: Bad sector found in cluster 140274210
Read Error: Bad sector found in cluster 140540182
Read Error: Bad sector found in cluster 140540202
Read Error: Bad sector found in cluster 140540221
Read Error: Bad sector found in cluster 140540770
Read Error: Bad sector found in cluster 140540874
Read Error: Bad sector found in cluster 140540900
Read Error: Bad sector found in cluster 169811583
Read Error: Bad sector found in cluster 169886784
Read Error: Bad sector found in cluster 169886805
Read Error: Bad sector found in cluster 169886825
Read Error: Bad sector found in cluster 169894476
Read Error: Bad sector found in cluster 188272533
Read Error: Bad sector found in cluster 188272563
Read Error: Bad sector found in cluster 193095105
Read Error: Bad sector found in cluster 193095877
Read Error: Bad sector found in cluster 193095885
Read Error: Bad sector found in cluster 193095891
Read Error: Bad sector found in cluster 193095896
Read Error: Bad sector found in cluster 193095900
Read Error: Bad sector found in cluster 193095903
Read Error: Bad sector found in cluster 193095905
Read Error: Bad sector found in cluster 193095906

 

[USAFRet]

If Macrium is reading a bad sector, there is no means to force Macrium to see that as a valid sector.

Does any other clone/imaging tool work?

 

[Darkmatterx]

That's the only thing I didn't think to try. Every program I tried was to look for bad sectors or any sort of indication of a problem with the drive.

I guess I'll try Acronis, as I've seen it mentioned several times in this forum. Unless you have a different suggestion for me.


While I'm at it, I'd love to know which sites are still "safe" to go to for software reviews. A lot of sites these days are either, clearly put out by one software company or another, or appears to be nothing more then a paid for review site.

Hardware gets a lot more coverage, which makes it much easier for the community to weed out the shills. :/

Thanks!

Edit: Oh, I don't need to uninstall Macrium first do I? Stopping the service and preventing it from doing it's regular backups should be enough to prevent complications, correct?

 

[LinuxDevice]

Do you have Linux available? The dd tool is quite good about actually testing for bad sectors, although read-only (a write and read-back test can show more). If you are at least familiar with Linux, then even if you don't have Linux installed you could use a Live DVD (or thumb drive) type installer which will run entirely in RAM and doesn't actually install. The dd program would normally be used to write a disk directly to another disk of the same size or larger, but you can test via writing to /dev/null.

You could find the device via a command like "lsblk -f" and then use dd to output either individual partitions or the disk as a whole. An example of this without logging, where the "..." would need to be replaced by the disk or partition name, and instead of an output disk you could use "/dev/null:
dd if=/dev/... of=/dev/... bs=512 conv=sync,noerror status=progress

You could write a log of this (useful because it can take hours or days depending on read speed and disk size):
dd if=/dev/... of=/dev/null bs=512 conv=sync,noerror status=progress 2>&1 | tee log.txt

I like the KDE live DVD and thumb drive distributions, e.g., from Ubuntu or Debian.

While there you could also check the smartdrive info, e.g., if your drive has designation "/dev/nvme0n1":
sudo smartctl -a /dev/nvme0n1

Then when you reboot with the live DVD or thumb drive removed there is no Linux left on the system and nothing changed.

 

[USAFRet]

Edit: Oh, I don't need to uninstall Macrium first do I? Stopping the service and preventing it from doing it's regular backups should be enough to prevent complications, correct?

Correct.
Just stop the schedule.

 

[Darkmatterx]

Okay, sorry about that. I had a life thing, but then my system drive must have got angry due to being left alone, without enough attention being given to it, as today Macrium Reflect told me that my MFT is corrupted!

I actually had collected some information, although some of it conflicting, using different programs, both from the manufacturer, and 3rd party, but never got a chance to post any of it.

[30min later] OK, I looked, and I can't find where I stuffed the info. FML...

To sum it up, Reflect saw 1 or 2 bad sectors quite a few times, and once or twice it said I had between 3 and 6 bad sectors. I didn't like any of this, but chkdsk, crystaldiskinfo, and even Samsung's own software said everything was fine.

Another concerning thing that also had signs of being a past problem was that I only had 44% of my spare reserve SSD space left. I checked my Event Viewer logs and noticed that it has been that way since around Oct 2024, which means whatever was going on, it hadn't changed in quite a while.
[Edit:] I found a log I must have manually saved the text of, and in Aug of 2024 I had 51% reserve left.

In the end, after much confusion over conflicting data, 1 sector got fixed and I thought I was done with that. Reflect wasn't giving me an error for full backups anymore, and life was chaos, while my SSD was apparently good. Not.

The exact same bad sector reappeared in Reflects logs the other day. I did a read only chkdsk, and it told me that the volume bitmap is incorrect, aka MFT corrupt. OK, so I do a reboot chkdsk, but guess what? No log was saved! I checked Event Viewer right after the damn system came back up, and used the log filter but I found nothing. Foolish little me didn't watch the whole chkdsk scan so I had no idea what it did. Nothing came up in a repeat scan, so I figured it was fixed. That was on the 6th.

Today, as I said above, Reflect told me my MFT was corrupt, and chkdsk read only also says that my MFT bitmap is incorrect. All this, but no bad sectors...

I think I am done with this particular EVO... If it wants me to retire it that badly, I guess I will. /sigh

I have a chkdsk /r scheduled for next reboot, but before I do that, does anyone have any suggestions before I do that? I have images of the drive, so I'm fine that way, but daaamn why am I getting so much different information? The SMART data hasn't changed, but whether the drive "has a problem" or not seems to entirely depend on the program I use to check it. I can't remember when I bought this thing, but it's absolutely not that old! Hell, my previous PC's SATA 500GB SSD is in this PC now and it's still running like a champ. In case you're wondering, there is a rather nice metal heatsink over the M.2 problem child.

TL/DR: I wrote a novel because simply typing out, FML, SMH, wasn't going to cut it...
Also, you're a GD champ if you actually read all of that.

One last question before I wait a bit to see if I luck out with a quick reply, and then just F it and reboot to chkdsk. Is there any way to get Windows to give me a notification when it finds a bad sector and grabs some of my reserve space? My potato brain is too tired to remember to check this <MOd Edit> stuff every few weeks.

Thanks.

DM

 

[USAFRet]

I have a chkdsk /r scheduled for next reboot, but before I do that, does anyone have any suggestions before I do that? I have images of the drive, so I'm fine that way, but daaamn why am I getting so much different information?

If you have full drive images, let it go.

If it is still under warranty, free replacement.
If not under warranty....it had a good life.


Drives die.
Eventually, all of them.

 

[Darkmatterx]

Fair,

Champ.

 

[Darkmatterx]

Ended up getting a migraine last night so haven't done anything yet. Stupid weather. 🙁

No more warranty.

Since I'm going to be placing an order, let me ask you a couple quick(er) questions.

I could use a suggestion for a replacement drive. I do have 2 M.2 slots, and I hear that some people try to get above blistering, blistering speeds by running them in RAID 0 (if the MB is capable of that) but I have no idea if that is actually worth it in any meaningful sense these days.

I know that your position on having a backup slot will likely fall towards RAID backup, NOT RAID 0, but since I do take full images or differentials daily, that's less of a concern for me.

Obviously, speed is important, but if it has the reliability rating of a rotten potato, obviously there's no point to it. I don't really have a set price point ATM. Nothing cheap or below 1TB, but I'm not selling my car for a new SSD either. Somewhere in between that. heh

Let me throw out some of my system's specs. I may need to update the BIOS depending on the drive recommendation, so I need to have a USB ready for that before I do anything.

Lastly, I've been using the heatsink's that came with the MB for the M.2 drive. The spec's are below, but here's a photo of the board. Do you recommend I replace them with a 3rd party version?

"The specs" & an image of the board so you can see the heatsink's it comes with.
[I tried uploading and embedding the image, but it doesn't like Imgur I guess.]

https://imgur.com/a/YZTu2bf

View: https://imgur.com/a/YZTu2bf


OS Name Microsoft Windows 10 Pro
Version 10.0.19045 Build 19045

System Model X570 AORUS PRO WIFI
System Type x64-based PC
System SKU Default string
Processor AMD Ryzen 7 5800X 8-Core Processor, 3801 Mhz, 8 Core(s), 16 Logical Processor(s)
BIOS Version/Date American Megatrends International, LLC. F36c, 2022-05-12
SMBIOS Version 3.3
Embedded Controller Version 255.255
Installed Physical Memory (RAM) 64.0 GB
BIOS Mode UEFI

 

[USAFRet]

and I hear that some people try to get above blistering, blistering speeds by running them in RAID 0 (if the MB is capable of that) but I have no idea if that is actually worth it in any meaningful sense these days.

Do not do this.

Those "blistering speeds" may only be seen in artificial benchmarks.

In actual user facing use? No.

I know that your position on having a backup slot will likely fall towards RAID backup,

No.
RAID, of any type, is not a backup.
You're probably thinking of RAID 1? No.

Regular automated backups are your friend.

 

[Darkmatterx]

OK thanks.

Do you have a brand/model suggestion?

 

[USAFRet]

OK thanks.

Do you have a brand/model suggestion?

Of what? SSD?

Samsung, WD, Crucial....

 

[Darkmatterx]

[Edit 2:] WHY!? Why is nothing ever easy with me!?

Thanks! But... curve ball. It looks like it is still under warranty. My lordy! lol

[Edit:] As I was typing this out, I remembered curve ball #2 & 3. SMART data thresholds & odd inclusion of D drive recovery partition, which is 450MB.

---

Warranty is, 5 years or 600 TBW. Got it Nov 2020 and it only has 57.7 TBW. Just a bit below 600...

The problem I am likely to run into is that nothing has reached even caution status for the SSD in the SMART menu for Samsung Magician. Are they going to care that I keep seeing bad clusters that mess up my system? Will they care that I'm down to 44% provisioning left despite only having written 57.7 TB of data to the drive?

Remember, I'm having bad sectors pop up on a regular basis, but the critical threshold in Samsung Magician is 10%, while I'm at 44%. I also don't understand the discrepancy between Reflect and chkdisk. Sometimes they agree, and sometimes they don't.

Guess I'll be contacting them and seeing what sort of CS the company has. If I am buying a new drive instead of warranty, it def won't be Samsung!

BTW, in this particular case, again, conflicting data, chkdsk said there were no bad sectors despite what Reflect says, and the corruption in my MFT was this,

CHKDSK discovered free space marked as allocated in the
master file table (MFT) bitmap.
Correcting errors in the Volume Bitmap.

Probably not the worst MFT problem that can appear, but it probably shouldn't have happened in the first place. Macrium has no support for their free version, so unless you guys here have any thoughts, I'm not sure how to figure out what is going on here. I also still don't know why Reflect is telling me that I need to backup the reserve partition on my D drive, as well as all of C.

I'm guessing it's an error due to the D drive being my C drive in my past system, but I'd have to disconnect my D drive and boot up my system just to see what would happen. Even then, I'd probably want to do a full backup image (which I'll do) and then another with that 1 D drive partition not selected, and see if it actually restored Windows.

[Edit 3:] 3rd edit. /sigh. I'm doing a full sector by sector backup right now to see if Reflect still sees any bad sectors. For actual replacement purposes I'm guessing I'll want to do only an intelligent sector backup.

FML SMH - I may have that put on my tomb stone... Frustration level increasing rapidly.

 

[USAFRet]

Use it (as a secondary device) until it dies.
Replace it when it does.

 

[Darkmatterx]

That's not a bad idea. The whole reason I have that D drive SATA SSD in my PC is for when I do a bit of video editing. Ironically, that old SATA SSD from Samsung has had zero problems and I have no clue how old it is.

I guess I'll still contact Samsung CS, since the provisioning and errors are pretty extreme for the amount of TB written. I'm not expecting much, but it never hurts to ask.

Oh, right. Did you have a 3rd party heatsink suggestion for an M2 drive? I'm wondering if this was a heat buildup issue.

I think I used the M2 slot that the board considers to be #1, but that may have been a bad call, as that's closest to the graphics card.

Thanks again. Last post, I probably, hopefully, promise.

 

[USAFRet]

That's not a bad idea. The whole reason I have that D drive SATA SSD in my PC is for when I do a bit of video editing. Ironically, that old SATA SSD from Samsung has had zero problems and I have no clue how old it is.

I guess I'll still contact Samsung CS, since the provisioning and errors are pretty extreme for the amount of TB written. I'm not expecting much, but it never hurts to ask.

Oh, right. Did you have a 3rd party heatsink suggestion for an M2 drive? I'm wondering if this was a heat buildup issue.

I think I used the M2 slot that the board considers to be #1, but that may have been a bad call, as that's closest to the graphics card.

Thanks again. Last post, I probably, hopefully, promise.

I don't have any specific heatsink recommendations.