Question A lot of MFT corruption and orphaned files

[boondocks]

Hi guys. As the title says, this is becoming problematic, and I'm running chkdsk every day now.
I have an LSI 9311 card and a Lenovo expander board.
Current running 11 HDD's and 4 SSD's, with an NVME boot drive.
I've lost all data on two HDD's over the last two months with apparently corrupted MFT's.
Running Windows 11 Pro, and the boot BIOS on the LSI card is disabled.
Any clues?

 

[Shmee]

Have you checked the health / SMART on your drives yet?

 

[boondocks]

Have you checked the health / SMART on your drives yet?

Yes. All looks good.

 

[fzabkar]

Perhaps the common factor is RAM, either on the motherboard or on the HBA. Are you able to disable write caching?

Can you upload a CHKDSK log?

 

[boondocks]

Perhaps the common factor is RAM, either on the motherboard or on the HBA. Are you able to disable write caching?

Can you upload a CHKDSK log?

I don't have any saved ones. I did notice in event viewer for yesterday 3 drives were "surprised removed". Don't think I've ever seen that before. None of the 3 have gone offline...
All the HDD's have write caching disabled as of now....see how it goes.

Thanks for your response.

 

[boondocks]

Wow just noticed that yesterday "the driver detected a controller error on \Device\RaidPort0
The source is LSI_SAS2i. Not sure what that means exactly. LSI card (8i)? Lenovo expander card? (Yeah it says LSI but it's ahead of my game)
Little lost here....

 

[Shmee]

Hmm yeah sounds like something may be up with your LSI card. Though since someone mentioned RAM, you could always run a memtest, as faulty RAM can cause corruption.

 

[boondocks]

My BIOS has memtest built in and I ran it on the system DRAM. Passed.
Just today I found this same drive for the third time has a corrupted MFT.
I have swapped this drive to a different cable on the controller but no change.
I'm currently running a long selftest on the drive with SeaTools.

I've got an LSI 930016i coming. I've tried this card before with no success and returned it but that was before I figured out I had to put the card in the x16 slot, get things working, then I can swap it to another slot. What I had to do with the 9311 card. Apparently some of these Asus consumer boards take some patience.

 

[fzabkar]

What are the model numbers of the problematic HDDs? I recall a thread at HDD Guru where one particular model was affected by silent data corruption. I even reported it to Seagate's management, but they didn't give it serious consideration.

 

[boondocks]

What are the model numbers of the problematic HDDs? I recall a thread at HDD Guru where one particular model was affected by silent data corruption. I even reported it to Seagate's management, but they didn't give it serious consideration.

This is an MDD 18TB. Since MDD are rebranded drives, I don't know who made it. The model and serial numbers in this case do not match any Seagate drives as some MDD drives I have bought in the past.
I have returned the drive and the seller will send a new one upon receiving the one sent.

 

[boondocks]

I do also have a 10TB Seagate ST10000VN0008-2PJ103 with 25536 hours on it, that while not experiencing MFT corruption, is starting to require weekly checking with chkdsk as I guess some files need to be (my words) accounted for/indexed properly.
I believe this is an Iron Wolf Seagate drive.

 

[fzabkar]

This is an MDD 18TB. Since MDD are rebranded drives, I don't know who made it.

If the PCB part number starts with 0B or 0A, it's a HGST HDD. If it starts with G, it's Toshiba, 2060 is WD, and 10 is Seagate.

 

[mikeymikec]

Yes. All looks good.

What's the SMART CRC reading?

---

RAM has been mentioned, but it seems unlikely that RAM problems would only affect storage. I wonder if the RAID card has its own cache.

Is Windows shutting down correctly? Have a look at the system logs or reliability monitor and see if Windows is reporting anything along those lines.

 

[fzabkar]

https://forum.hddguru.com/viewtopic.php?f=1&t=40740

 

[Shmee]

https://forum.hddguru.com/viewtopic.php?f=1&t=40740

Care to summarize or explain what you think of the thread? Perhaps what I gather though, is that the common link seems to related to Seagate drives? Perhaps just issues with drives in some way, I don't usually use Seagate HDDs, generally I prefer quality WD or HGST drives. Generally higher end or data center 5 year warranty drives.

 

[fzabkar]

The user in that HDD Guru thread detected errors in certain files because the BTRFS file system appends checksums to data and metadata. His Seagate drive reported no errors in the SMART attributes, yet there were several "Reported Uncorrectable Errors" in the drive's Device Statistics (GP Log 0x04). These are two different components of the drive's firmware.

Therefore, if this same user had used NTFS rather than BTRFS, he would have been faced with insidious data corruption (because the drive would not have reported a read error to the OS).

 

[mikeymikec]

I think it wouldn't be a bad idea for @boondocks to post screenshots of the SMART stats.

 

[boondocks]

I have replaced a drive, and have been running chkdsk every day on all drives and not having any problems now. I really appreciate all the input.
Now I have switched from an LSI 93118i card to a 930016i and it's giving me fits. But I will start a new thread on that.