Thousands of parity errors

[Carpet3]

Hi,

 

I've just started getting tens of thousands of errors on my server.

 

First check got 46724, I did a subsequent (none correcting) check and it came up with 99800.

 

Could someone help with what's going on? Diagnostics attached before I rebooted it.

 

I can see the following error again and again in the log:

 

Mar  4 23:56:24 Serenity kernel: ata10: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Mar  4 23:56:24 Serenity kernel: ata10.00: configured for UDMA/133
Mar  4 23:56:24 Serenity kernel: ata10: EH complete
Mar  4 23:58:25 Serenity kernel: ata10: limiting SATA link speed to 3.0 Gbps
Mar  4 23:58:25 Serenity kernel: ata10.00: exception Emask 0x10 SAct 0x0 SErr 0x190002 action 0xe frozen
Mar  4 23:58:25 Serenity kernel: ata10.00: irq_stat 0x80400000, PHY RDY changed
Mar  4 23:58:25 Serenity kernel: ata10: SError: { RecovComm PHYRdyChg 10B8B Dispar }
Mar  4 23:58:25 Serenity kernel: ata10.00: failed command: READ DMA EXT
Mar  4 23:58:25 Serenity kernel: ata10.00: cmd 25/00:40:d8:c5:9b/00:01:05:00:00/e0 tag 20 dma 163840 in
Mar  4 23:58:25 Serenity kernel:         res 50/00:00:d8:c5:9b/00:00:05:00:00/e0 Emask 0x10 (ATA bus error)
Mar  4 23:58:25 Serenity kernel: ata10.00: status: { DRDY }
Mar  4 23:58:25 Serenity kernel: ata10: hard resetting link
Mar  4 23:58:31 Serenity kernel: ata10: SATA link up 3.0 Gbps (SStatus 123 SControl 320)
Mar  4 23:58:31 Serenity kernel: ata10.00: configured for UDMA/133
Mar  4 23:58:31 Serenity kernel: ata10: EH complete

I'm guessing it's either the card or one of the drives but I don't see how to identify which?

 

I recently (before these errors showed up) did a quick SMART test on all the drives and they all passed.

 

I know one of the disks (7) has a lot of UDMA errors but this was due to a faulty cable a while back and has been fine for a long time.

 

Thanks in advance

 

 

 

serenity-diagnostics-20200304-2331.zip

[Carpet3]

For reference, I'm using all the ports on my motherboard and a M1015/SAS9220-8i card in IT mode

[JorgeB]

You should start by running memtest, also check replace cables on parity2

[Carpet3]

Hey, thanks for the response.

 

Why do you say parity 2? How did you identify that was potentially the problem? 

 

I recently (a few weeks ago) upgraded that drive. It precleared ok

Edited March 5, 2020 by Carpet3

[JorgeB]

2 minutes ago, Carpet3 said:

Why do you say parity 2? How did you identify that was potentially the problem? 

You can see in the syslog that ata10 is parity2, doesn't look like a disk problem, looks like a connection problem.

[Carpet3]

Ah ok, I think I see how you determined that.

 

Thanks for the help. I'll try a memtest and change that cable over and see how it goes