If you get sth. like this (and best case your raid 1 goes crazy)
daywalker kernel: [ 498.324514] ata5.00: status: { DRDY }
[ 498.324516] ata5.00: failed command: WRITE FPDMA QUEUED
[ 498.324520] ata5.00: cmd 61/60:e8:e2:7d:0e/01:00:9c:00:00/40 tag 29 ncq 180224 out
[ 498.324520] res 50/00:d8:7a:87:0f/00:00:9c:00:00/40 Emask 0x10 (ATA bus error)
[ 498.324522] ata5.00: status: { DRDY }
[ 498.324524] ata5.00: failed command: WRITE FPDMA QUEUED
[ 498.324527] ata5.00: cmd 61/88:f0:4a:7f:0e/01:00:9c:00:00/40 tag 30 ncq 200704 out
[ 498.324528] res 50/00:d8:7a:87:0f/00:00:9c:00:00/40 Emask 0x10 (ATA bus error)
[ 498.324530] ata5.00: status: { DRDY }
[ 498.324533] ata5: hard resetting link
[ 498.644016] ata5: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 498.646630] ata5.00: configured for UDMA/133
[ 498.660052] ata5: EH complete
[ 499.755828] ata5.00: exception Emask 0x10 SAct 0x7ffff9ff SErr 0x400100 action 0x6 frozen
[ 499.755831] ata5.00: irq_stat 0x08000000, interface fatal error
[ 499.755834] ata5: SError: { UnrecovData Handshk }
[ 499.755836] ata5.00: failed command: WRITE FPDMA QUEUED
[ 499.755840] ata5.00: cmd 61/00:00:b2:51:ed/04:00:4d:00:00/40 tag 0 ncq 524288 out
[ 499.755840] res 50/00:98:ba:6a:ed/00:01:4d:00:00/40 Emask 0x10 (ATA bus error)
[ 499.755842] ata5.00: status: { DRDY }
[ 499.755844] ata5.00: failed command: WRITE FPDMA QUEUED
[ 499.755848] ata5.00: cmd 61/08:08:b2:55:ed/02:00:4d:00:00/40 tag 1 ncq 266240 out
[ 499.755848] res 50/00:98:ba:6a:ed/00:01:4d:00:00/40 Emask 0x10 (ATA bus error)
[ 499.755850] ata5.00: status: { DRDY }
you might want to check a different S-ATA cable first. (Chances for a kernel bug are very bad too…) Would have spared me a lot of time… If you have got the opinion that the cable is innocent because windows is working properly on the same hdd well guess again because windows easily ships around this mess by falling back into not using dma 🙂
(If another cable on another SATA connector does not change anything the hdd is probably broken)
UPDATE:
http://lime-technology.com/wiki/index.php?title=The_Analysis_of_Drive_Issues
is pretty informative though!
UPDATE II:
Had a short conversation with a computer shop owner the other day. He told me that the SATA cables provided (for free) with the mainboard are much more likely to fail than others. Don’t know if he just wanted me to buy more cables (i did not) but my faulty one was indeed provided with the mainboard and he didn’t knew that in advance.