On Mon, 01 Sep 2014 21:13:26 -0700 Linda Walsh wrote:
Nope... It really is the raw error read rate... It matches up exactly with the Hardware ECC correction rate. All of the raw errors are currently being corrected by ECC. When they aren't, you'll start seeing sector remapping, which is when you start to really notice disk slowdowns --- they prevent contiguous reads.
I don't know how long one can use ECC to get by with, BUT I've always been told it's the first step towards an unreadable sector that gets remapped.
Well, this technology has come a long way in 20 years. Here I've been, minding my own business, operating on the background assumption that S.M.A.R.T. _now_ is pretty much what it was back then -- maybe with more 'bells and whistles' accrued over that time. My understanding from when I delved deeply into it (I was working in the storage products industry, btw, in Silicon Valley) was that 'raw read error rate' and 'seek error rate' related to 'cache misses' not actual, physical read errors. My understanding was that these metrics were being used statistically, internally, over long periods of time and across a massive install base by the manufacturers to monitor and improve caching algorithms. In Seagate's own words*: "The idea ... is to predict a failure before it happens. Various attributes are being monitored and measured against certain threshold limits. If any one attribute exceeds a threshold then a general SMART Status test will change from Pass to Fail." and "The individual attributes and threshold values are proprietary and we do not offer a utility that will read out the values." Finally, "As a practical matter, the technology supporting SMART is constantly being improved. Each new design incorporates improvements that increase the accuracy of the SMART prediction. As a matter of policy, Seagate does not publish attributes and thresholds. Therefore, if you wish to test the drive for physical integrity, please use our SeaTools diagnostic software." From a practical standpoint, a 'fail' in lieu of a 'pass' with a drive that remains operational, under warranty, is sufficient to obtain a warranty replacement drive from Seagate. In the OP's specific case, if the attribute 'reallocated sector count' is correct for both drives, they're at zero (0). All things being equal, between that and SMART test 'pass' would give me a fairly high degree of confidence that failure doesn't seem to be imminent. I'd still, as I always do, back up my important data. (*) <http://knowledge.seagate.com/articles/en_US/FAQ/203971en> regards, Carl -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org