ERROR Fetching qualities with new hardware

[scoville07]

@madMAx43v3r and all

I recently moved my farmer and harvester to new hardware (new computer build), basically from pcie3 to pcie 4 and from Windows 10 Pro to Windows 11 Pro. The Nvidia 3060 I use on both is still the same. Still 64gb ddr4, CPU stlightly different (xeon vs. i5). Drivers still up to date.

The roughly 24000 C18 plots are still the same, the were procuded from a separate machine (3060ti, xeon, Windows 10).

Since I moved I get about 10% ERROR Fetching qualities on my plots. Does this correlate to the Hardware change of the farmer and harvester? Any quick guess? Do you think it makes sense to replot all the faulty plots?

Note, there is a difference in the Fetching qualities error message:

badbit or failbit after reading size
vs
Invalid file C:_chia\

Would it makes sense to replot all "invalid" ones?

Appreciate your feedback.

Regards
Christian

[madMAx43v3r]

Can you post full error messages?

[scoville07]

2024-02-20T00:25:39.234 harvester chia.harvester.harvester: ERROR Exception fetching qualities for C:_chia\xyz\plot-k32-c18-2024-01-23-12-56-xyz.plot: badbit or failbit after reading size 4 at position 62947377408 PlotInfo(prover=<chiapos.DiskProver object at 0x0000024AC037DB70>, pool_public_key=None, pool_contract_puzzle_hash=<bytes32: 867dde82ec3e97d90a93a65e436...>, plot_public_key=, file_size=64021215744, time_modified=1706017002.7199433)

vs.

2024-02-20T00:26:10.691 harvester chia.harvester.harvester: ERROR Exception fetching qualities for C:_chia\xyz\plot-k32-c18-2024-01-30-14-19-xyz.plot: Invalid file C:_chia\xyz\plot-k32-c18-2024-01-30-14-19-xyz.plot PlotInfo(prover=<chiapos.DiskProver object at 0x0000024ABECA6270>, pool_public_key=None, pool_contract_puzzle_hash=<bytes32: 867dde82ec3e97d90a93a65e...>, plot_public_key=, file_size=64045079216, time_modified=1706629562.8610246)

[madMAx43v3r]

Looks like a HDD issue, it fails to read the files.

[scoville07]

Thanks.

That said, as this error affects basically more than half of my 100 disks, I doubt it is a hardware issue. They are all attached with 10bay JBOD Chassis via USB (five on each, harvester and farmer, through PCIe USB cards). You think it is related to USB connection issues?

And in addition, would it makes sense to at least replot the "invalid" files?

[scoville07]

I just had a look in the event log of windows, not sure why I didn't do that in the past, probably because I did not really encouter issues and I probably never looked in the chia logs as well because of that. I just looked in the chia logs after the hardware swap.

Long story short, seems windows (surprise surprise) doesn't like that much of disks behind single USB ports and occasionally throws errors on disks with same system ID, at the same time the "fetching errors" occur. Although all disks are recognized one after another without issues (at least at the frontend ;))

So I have to live with that or swap to RAID or storage chassis, NAS, whatever... ;( Damn expensive... ;)

[madMAx43v3r]

And in addition, would it makes sense to at least replot the "invalid" files?

No, "invalid" in this case means there was a hardware / OS issue to open the file.

USB is tricky yes...