[Solved] Debugging UDMA CRC error count

[mattenz]

My UDMA CRC error count keeps going up on all drives, so I'm after some advice on how to start addressing it.

 

UNRAID is running on a HP DL320 G7, six 8TB SATA drives are mounted in a HPD2600, connected via a LSI SAS 9212-4i4e HBA.

[John_M]

UDMA CRC errors are SATA link errors caused by hardware problems, such as bad cables, connections, backplanes and sometimes power distribution. Less comon causes include bad controllers and bad drive electronics. If just one disk was affected you'd be advised to replace the sata cable and check that the power connector is securely plugged in. Since all your disks are affected it's going to be something common to them all, such as the connecting cable, the enclosure, its power supply or backplane, or the HBA.

[mattenz]

Thanks, I've remounted everything on the server side (PCI riser, HBA card, cable), and started a parity check, will see how it goes.

[JorgeB]

8 hours ago, mattenz said:

LSI SAS 9212-4i4e

There are also some LSI firmwares that have known issues with that, what firmware are you using? Or post diags.

[mattenz]

3 hours ago, JorgeB said:

There are also some LSI firmwares that have known issues with that, what firmware are you using? Or post diags.

Probably a bit of a rookie question, but how do I check the firmware version? Should I just be shooting for the latest version?

[JorgeB]

3 hours ago, JorgeB said:

Or post diags.

 

[mattenz]

Like the whole thing?

tower-diagnostics-20210513-2158.zip

[JorgeB]

LSISAS2008: FWVersion(20.00.02.00)

This is one of the problem firmware releases, update to 20.00.07.00.

[mattenz]

Just now, JorgeB said:

LSISAS2008: FWVersion(20.00.02.00)

This is one of the problem firmware releases, update to 20.00.07.00.

Thanks will do, is it worth letting the parity check run?

[JorgeB]

I would cancel and run it after the update.

[mattenz]

12 hours ago, JorgeB said:

I would cancel and run it after the update.

That seemed to remove the errors (at 3.0% parity check with no errors). Also gave me a chance to learn how to mount from the DL320 SAS controller! Thanks again!