I have had my d5400xs up and running for nearly two years ... no problems until last month when one of my bios striped velociraptors (2x) started showing SMART error code on the raid boot status screen.
I promptly backed up and downloaded and installed the latest bios, and also drivers for Windows 7.
1) there is apparently no way to read the SMART codes to know what was going on. The bios does not tell, the raid console for windows only reports the existence of a problem and provides a "reset status" link. and none of the SMART utilities for windows reveal any problem.
2) got another velociraptor and used "dd" on a linux box to copy the drive with the status codes to the new drive. THIS DID NOT WORK. Even with identical bits on both drives, the BIOS failed to recognize the new drive as a member of the raid array ... what gives with that? is the raid bios using some offline parameter block in the drive's nvram?? is it using maintenance sectors not accessible to a raw storage driver?
3) finally gave up on rebuilding the raid array and simply told the bios to create a new raid array from the non-failing drive and the new drive, then restored my backup. This did not work either ... now both drives are showing SMART codes at boot ==> hence my conclusion that the IO controller is failing. All three velociraptors passed a 5 hour stress test on another system with ZERO issues.
4) the iastor.sys driver is a POS. the SMART error reports often just hang the system to the point where only a hard reset or a power-cycle work. removing the Intel Raid manager software and the iastor driver clear the system hangs.
the bottom line questions: is there a bright line diagnostic utility which will show the actual SMART codes for the BIOS managed raid array? is there a definitive way to prove that the MB raid controller chipset is what is actually getting flaky?