UDMA CRC error counts / I/O error, dev sdu, sector 3907029160 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

[shire]

Hello,

 

after I replaced a drive (one with reallocated sectors) I did a rebuild. This worked without any problems. I also decided to check on the SMART status of my other drives. Two are problematic. Both have ‘UDMA CRC error counts’ in the 4xx (/sdu, /sdl).  Besides that, drive /sdu wrote some errors into the logfile while rebooting.

 

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 3907029160 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 3907029104 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 3907028176 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 3907028216 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 3907026080 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 32 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 2048 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 120 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2

Aug 29 13:11:11 Tower kernel: I/O error, dev sdu, sector 264 op 0x0:(READ) flags 0x80700 phys_seg 29 prio class 2

 

I did an extended SMART test on both drives afterwards (passed both). Could anyone check the logfiles? I have another new drive ordered that will arrive today. I would like to decide on which one to replace next.

 

I will check the cabeling on both drives. I have also the complete diagnostic if needed.

 

Bye.

WDC_WD20EARS-60MVWB0_WD-WCAZA4043525-20230830-1012.txt WDC_WD20EARS-22MVWB0_WD-WCAZA6163992-20230830-0924.txt

[itimpi]

3 minutes ago, shire said:

Both have ‘UDMA CRC error counts

These are not normally an issue unless they are increasing at a steady rate.    They indication connection issues between the drive and the host machine and are typically caused by power or SATA cabling issues.   If one is encountered then the host will retry the transfer so the typical impact is just to slow things down unless the retries all fail and as a result a transfer actually fails.  It is worth noting that the CRC count never resets to 0 so once you have some if you can stop them increasing you have fixed the issue.

[shire]

The I/O errors happened during the last reboot. I rebooted again and they did not reappear.

 

So...I gues I have to observe both 'problems'...

 

Thanks...

[Frank1940]

2 hours ago, shire said:

So...I gues I have to observe both 'problems'...

 

If I remember correctly, on the GUI Dashboard, if you click on the 'problem' symbol, you are presented with an option to 'accept' the problem and things will go 'green' unless another CRC error occurs. 

[shire]

5 hours ago, Frank1940 said:

 

If I remember correctly, on the GUI Dashboard, if you click on the 'problem' symbol, you are presented with an option to 'accept' the problem and things will go 'green' unless another CRC error occurs. 

Yes..those UDMA CRC errors cause an error in the drive overwiev. I acknowledged them and they are back to normal (green).