Contradiction in results provided by Verify feature
Performed verify feature on 2 different drives and 2 different Macs and obtained the following contradiction in the results of these 2 drives. In BOTH cases the following dialogue box is presented at the end of the verify process:
BUT the SoftRAID log file states:
for one drive:
Oct 18 0119 - SoftRAID Application: The verify disk command for disk disk0, Label: "Drive 1", SoftRAID ID: 06A58D4B3BA46900, SATA bus 0, id 0 hung while reading (offset 5,000,969,322,496, i/o block size = 11,755,520). This disk should be replaced immediately.
Oct 18 0125 - SoftRAID Application: The verify disk command for disk disk0, Label: "Drive 1", SoftRAID ID: 06A58D4B3BA46900, SATA bus 0, id 0 completed successfully.
Oct 15 1518 - SoftRAID Application: Verifying the disk disk170, Label: "Drive 3-bad", SoftRAID ID: 08564E1BACCC7980, USB3 bus 0, id 0. All the sectors on this disk will be read to ensure that they are reliable.
Oct 16 0037 - SoftRAID Application: The verify disk command for disk disk170, Label: "Drive 3-bad", SoftRAID ID: 08564E1BACCC7980, USB3 bus 0, id 0 hung while reading (offset 6,001,174,511,616, i/o block size = 614,400). This disk should be replaced immediately.
Oct 16 0043 - SoftRAID Application: The verify disk command for disk disk170, Label: "Drive 3-bad", SoftRAID ID: 08564E1BACCC7980, USB3 bus 0, id 0 completed successfully.
This is an obvious contradiction.
Tests were conducted multiple times to confirm results on two different computers.
Seems like something needs to be corrected.
This is a bug in the drive reporting. We are going to resolve this by working around it, either by suppressing the error, or finding a way to not complete the last write to the disk.
Note that in both your cases, this is the last sector on the disk.
There is a disk on the engineering desk that reproduces this issue, so hopefully it will be sooner than later!
Thanks though, good feedback!
So does this mean that the drive is NOT faulty if it presents with this error in the log? I am in the process of dealing with an RMA for one drive and preparing myself for the other.
Note that with one drive connected to the bus I do experience stalling intermittently and unpredictably... so am I a little confused if the drive is truly the problem or not since SoftRAID is not always great at isolating and confirming WHICH drive might be the cause of problems. In my case, both drives are inside an iMac so not easy to disconnect one just to confirm.... the only result pointing to this drive is the Verify test.
It is NOT faulty, no, if the verify error is the last sector on the disk. You can look in the disk tile and if the error is within 128K of the end of the disk (the typical seek size), then it is another issue.
SoftRAID cannot detect SMART if you have a fusion drive setup, if you look in the disk tiles, you will either see SMART passed, or a -, depending. If a drive is stalling, AND SoftRAID cannot get SMART on the drive, try a third party tool like DriveDx (trial mode is OK) and make sure the disk does not have a pending failure.
BTW, I have 3 identical 6TB drives (one is the RMA replacement)... only one reports this problem (on top of 5TB). No Fusion drives involved. SMART tests all show "passed".
When you say "This is a bug in the drive reporting." is this a DRIVE reporting the problem or SoftRAID misreading what the drive is reporting, or some such thing?
I am not sure, it appears to be a discrepancy in how many sectors the drive has vs. reported. Some drive models have this issue across the board. We will need to work around it, as it does not appear to be a fault, but a reporting issue.
So I do not need to worry about my 5TB drive and I do not need to replace it? It is still under warranty for a few more months.
Also, why would the Verify feature give this result but this does not show up with the Certify feature?
I do not know. When you certify, it sometimes can fail on such drives, but only on the last pass, which is also weird. The key is whether it is on the last read on a disk. That is the clue it is not a "real" issue with the drive.
ok thx... strange scenarios.
BTW, here is the SMART report from TechTool Pro for the 5TB drive:
So not sure what to make of this when combined with the SoftRAID results... though in that report I see issues identified with the Spin-up time... among other issues.
What we know is 196 197 198 are primarily associated with disk failure.
If you have any of those symptoms, SoftRAID will show the disk as predicted to fail ((look in the disk tiles, SMART should show passed and the 5TB should theoretically show reallocated sector count, 0 or higher. If a -, then SMART is unavailable on the disk for us (USB?)
Yes, SMART shows passed and reallocated sector count = 0 for 5TB drive. FYI it is on internal SATA bus.
I know the chart is not from your software but any idea what 197 & 198 showing as halfway between Pass & Fail means? And what the Spin-up Time showing near Fail means (again I get stalling on the SATA bus...just cannot confirm if it is this drive or the SSD on the same internal SATA bus).
The 197 198 should be kind of pass fail, there are either symptoms or not. I do not know of any middle state, being in the middle implies a low number is OK, where the various disk studies show ANY reallocated sectors are a sign of failing hene pending reallocations are a sign of failure also, unless something like a certify clears the state.
Spinup time, i do not know. We do know there is no clear statistical correlation with failure from for instance backblaze. So I do not know whether this is important or not.
Thanks for the feedback. I'll have to dig into this a little further...
But to clarify, if SoftRAID is reporting a SMART pass but TechTool Pro is suggesting SOME reallocated sectors, using SoftRAID Certify could potentially clear the state if it finds nothing? Are SMART parameters updated by Certify?
SMART parameters are very clear. Either they say zero, or they have a number. These are reported by the drive.
Certify can clear "unreliable sector", or sometimes "pending reallocation" counts, but generally it will trigger the drive to reallocate them. Once a sector is reallocated, that is permanent. But there is no in between state, these are numbers reported by the drives.
I do not know what TechTool is seeing in this case, if it does not give a specific number.