I'm currently swapping my 3TB disks to 4TB now.

I've successfully preclear all the 4TB disks, without any issue.

Now I'm doing the swap. And I've an error while rebuilding.

 

Mar 29 15:57:08 Tour kernel: ata4.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 (Errors)

Mar 29 15:57:08 Tour kernel: ata4.00: BMDMA stat 0x6 (Drive related)

Mar 29 15:57:08 Tour kernel: ata4.00: failed command: WRITE DMA EXT (Minor Issues)

Mar 29 15:57:08 Tour kernel: ata4.00: cmd 35/00:a0:60:f6:15/00:02:10:01:00/e0 tag 0 dma 344064 out (Drive related)

Mar 29 15:57:08 Tour kernel:          res 51/84:21:df:f7:15/84:01:10:01:00/e0 Emask 0x30 (host bus error) (Errors)

Mar 29 15:57:08 Tour kernel: ata4.00: status: { DRDY ERR } (Drive related)

Mar 29 15:57:08 Tour kernel: ata4.00: error: { ICRC ABRT } (Errors)

Mar 29 15:57:08 Tour kernel: ata4: soft resetting link (Minor Issues)

Mar 29 15:57:08 Tour kernel: ata4.00: configured for UDMA/133 (Drive related)

Mar 29 15:57:08 Tour kernel: ata4.01: configured for UDMA/133 (Drive related)

Mar 29 15:57:08 Tour kernel: ata4: EH complete (Drive related)

 

It seems that it's not something I may ignore, but I don't know on which drive this error is related, if it's about the interface or the disk itself, if I have to go back to the previous situation (perfectly possible, I've the copy of the flash and the 3TB itself), if it may help...

 

Thanks for any advice, ...

 

 

Check all the SATA cables, one is probably loose. (Its hard to tell which with the info provided), and if they aren't  LOCKING SATA cables, consider replacing them.

 

If it doesn't fix the problem, try and post the FULL syslog.

See this:

http://lime-technology.com/forum/index.php?topic=9880.0

or:

http://lime-technology.com/wiki/index.php/Troubleshooting#Capturing_your_syslog

In fact you're right. There is the syslog in attachment.

 

I've re-check twice every SATA connection on both side, and power connection too, but everything looks OK.

Now I've reboot the server, and started a parity-check.

Of course there is no error in the new syslog until now (the syslog below is the syslog WITH the error, of course.

How can I have a clue about the previous disk that could be the root of the problem in the syslog?

 

Thanks

syslog-2014-03-29_after_3Tb-4TB_swap_.txt

How can I have a clue about the previous disk that could be the root of the problem in the syslog?

The disks are assigned addresses (like 'ata4') in the order in which they are spun up and found.

So the assignment of a 'disk' to an 'address#' can change from boot to boot.

That's why a complete syslog is needed...the beginning of the log shows the physical disk to logical disk assignments.

 

See how things go. If there's no recurrence, don't worry about it...it was probably just a cable that needed to be jiggled into place. The next time you open up your array, you might think about swapping in locking SATA cables.

OK.

After some investigation, and following your advice, I guess I've found on which drive the ata4.00 is connected :

Mar 29 10:28:41 Tour kernel: ata4.00: ATA-8: ST4000DM000-1F2168,            Z300GAQB, CC52, max UDMA/133

and yes, in fact, it was the drive I've just swapped.

 

End of the parity-check, seems nothing I've to worry about.

I've found some locking SATA cable brand new in a drawer, I'll put those on the next swap (tomorrow?)

 

Thanks

 

Cheers! Glad you found it!