Quote:
Originally Posted by catkin
Yes -- an AMD SB600.
|
The SB600/700 AHCI support does have some issues. The significant
PMP bug was fixed in 2008; it causes a stall for a few seconds, but corrects itself. This one is easy to identify, look for "HW BUG" nearby in the logs.
The other issue I'm aware of (and assume catkin too is seeing) seems to be completely harmless. I've seen no data errors in 18000 hours of RAID-0 and RAID-1 use on a Gigabyte GA-MA78G-DS3H and two Samsung HD103UJ drives, over a number of kernels (mostly vanilla from kernel.org, for bleeding-edge radeon support -- RS780 on that MB).
This means that the "softreset failed" message itself in the logs does not tell much on SB600/700/etc. chipsets, as it is typically harmless. It does not mean it is
always harmless, as there may be a real problem behind it -- something like a badly seated SATA or SATA power cable, or a faulty SATA chip, for example.
Quote:
Originally Posted by grob115
Sorry which part of the "smartctl -a /dev/sda1" output should I focus on to find the number of reallocated sectors?
|
It is the raw value of ID 5, 'Reallocated_Sector_Ct'. Note that you may have to run the offline test to update the attributes. (The offline test should take less than a second to run.)
The value itself is not that important, as long as it's fairly small, but when it starts growing, the drive is dying.
Personally, I use scripts (with smartd disabled) to monitor both the drive temperatures and attributes, and only rarely run a self-test.