Byte-for-byte failure VSS+PHY, but no visible disk issue?

[timg11]

No RAID -three normal SATA drives. One SSD, two mechanical.

I tried backing up to a server drive instead of a local drive, but I still get the errors.

[timg11]

Interesting observation using FINDLBAF.

When I run this on my Samsung 840 SSD (The one that reports the validation errors), it reports "Found Partition: MBR 0 238473MB (0001) HPFS/NTFS"

When I run FINDLBAF on another drive (not the SSD), it reports "Found Partition: H-MARS-1.5T 1430796MB (0003) HPFS/NTFS"

"H-MARS-1.5T" is the volume name of the other drive. The volume name of the C: drive is "C-MARS-W7-ssd", but FINDLBAF reports it as "MBR 0". Why?

[TeraByte Support]

make sure you're running it on the correct drive number - right click
computer, manage, disk management, look at what drive 0 is, drive 1, etc..
that is the number to use with findlbaf


"timg11" wrote in message news:[email protected]...

Interesting observation using FINDLBAF.

When I run this on my Samsung 840 SSD (The one that reports the validation
errors), it reports "Found Partition: MBR 0 238473MB (0001)
HPFS/NTFS"

When I run FINDLBAF on another drive (not the SSD), it reports "Found
Partition: H-MARS-1.5T 1430796MB (0003) HPFS/NTFS"

"H-MARS-1.5T" is the volume name of the other drive. The volume name of the
C: drive is "C-MARS-W7-ssd", but FINDLBAF reports it as "MBR 0". Why?

[TeraByte Support]

different sectors at different times? check upperfilters to see if anything
closer to the drive than phylock that may update things, but have you tried
replacing the system memory ?

"timg11" wrote in message news:[email protected]...

No RAID -three normal SATA drives. One SSD, two mechanical.

I tried backing up to a server drive instead of a local drive, but I still
get the errors.

[timg11]

Yes, different sectors for each run. Byte-for-byte failure at LBA: 6780450, 245016703, 181786143, 100159223, 6249215, 248732191, 249862559, 106884465, 3323410, 172871717, 148404307

Please explain more about Upperfilters. There are many instances of that string in the registry. At HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{4D36E967-E325-11CE-BFC1-08002BE10318} (Class DiskDrive), I find upperfilters is REG_MULTI_SZ, and the value is "phylock PartMgr"


"

[TeraByte Support]

that looks fine.

have you replaced your memory and ensured the BIOS timings are correct for
the memory.

or try replacing the drive or moving it to a different cable/controller.


"timg11" wrote in message news:[email protected]...

Yes, different sectors for each run. Byte-for-byte failure at LBA: 6780450,
245016703, 181786143, 100159223, 6249215, 248732191, 249862559, 106884465,
3323410, 172871717, 148404307

Please explain more about Upperfilters. There are many instances of that
string in the registry. At
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{4D36E967-E325-11CE-BFC1-08002BE10318}
(Class DiskDrive), I find upperfilters is REG_MULTI_SZ, and the value is
"phylock PartMgr"


"

[timg11]

Memtest 5.0.1 detects no memory problems. This system is 5 years old and has been rock solid the whole time. BIOS timings have not changed, and are not overclocked or pushed beyond spec. Granted memory can age and fail. I have ordered replacement memory just in case.

The Samsung 840 SSD is new. Granted it could be defective in some strange way, but it seems odd that everything except the backup verification seems to be fine.

Trying other SATA ports is the next step.

One other point - my Gigabyte GA-P35-DS4 V2 motherboard has only SATA 2 (3 Gbps ports), so that's what I'm using. The SSD supports SATA 3, but in theory it is backward compatible.

[timg11]

I tried a different SATA port on the motherboard. I used a different SATA cable just in case.

No changes in the failures.

[timg11]

New memory installed - re-ran backup - same byte for byte verification failure.

[timg11]

I went back to IFW 2.82a just as a check, and there is a difference.

With over 10 backups attempted with IFW V 2.85, every "data as backed up" is a block of zeros, and "Data now on disk" is a block of mixed data.

With 2.82a, I still get a Byte-for-byte failure, but the difference is a single bit. Both blocks are mixed data.

Data As Backed up, row 29: "50 8B CB E8 C0 8C FF FF 8B 45 AC 8B 88 9C 00 00 "
Data Now on Disk, row 29: "50 8B CB E8 C0 88 FF FF 8B 45 AC 8B 88 9C 00 00"

One bit in one byte changed - 8C to 88, all other bytes are identical.

Next step is to order a new SSD drive. I'll try Toshiba this time.