Hi, I have an X-25M as my system/boot drive in a 2009 MacPro and lately I've been getting input output errors when backing up. There's been no issues re useability/performance that I can detect, is this a cause for concern? Is the drive in danger of failing? Many thanks. -dave
Well this isn't looking good. It's reporting 100 "uncorrectable errors." I think I'm outside of warranty too. Now I'm wondering when I replace the drive will my cloned backups be any good or will I have to do a major re-install of everything.
I really wonder about those SMART values, how could you have a "Power On Hours" of 30? Also, the "Time Occurred" is 30, that does not make sense. It is also odd that the errors are happening on reads, since when SSDs wear out, it's the writes that fail, not the reads. But your SSD could be failing, not wearing out.
IMO, SMART "information" can't be trusted with SSDs. For example, look at Duckie's SMART data. An SSD with 3.4 TB of writes has a Power-On Hours Count of 30? What a coincidence, so does darob's! My 80GB G2 has over 900 hours Power-On Hours, less than 1TB host writes, and is less than a year old. But who's data can we trust, if I claim it can't be trusted? This statement is from an article about SMART:
"As a result, S.M.A.R.T. is not always implemented correctly on many computer platforms, due to the absence of industry-wide software and hardware standards for S.M.A.R.T. data interchange."
The error count on both of the 80GB G2 SSDs I have on my PC is 100. One is an OS drive about 1/3 full, the other new with 200MB of files on it. How could we possibly have the identical number of errors, and that number also by chance is a nice and clean 100?
I'm sorry I'm not helping you with your problem, my point is that IMO we should not assume that SMART data is actually meaningful, unfortunately.
I was wondering about those obviously wrong numbers, and whether input output errors are even related to the smart status and not something else entirely. I don't know what's going on under the hood here I'll just make sure I have another drive ready to go and dig up some pre-error backups. Thanks for the replies.
You are welcome. Rather than blame the SSD, I should say that the SMART reporting itself, from the data sent by the device, to it's interpretation by... whatever, including ourselves, is either confused or not done correctly.
For example, many SMART data values have a range of 1 to 253, 1 representing the worst situation, and 253 the best. Then, the device manufacture chooses what the "normal" value is, usually 100 or 200. So it seems that we need to be smart enough to interpret the SMART data correctly, but how do we know what values mean to a particular device?
In Apple-land (orchard?) many of their PCs use ECC memory (error correction) which is not used in the MS PC world. Perhaps at times there are errors that cannot be corrected, which is noted or logged as you are seeing. Whether or not that is significant, I don't know. So you may be seeing a problem of some sort, although the particular number or SMART value does not indicate exactly what that is, or could be a count of errors, as you thought. So overall the SMART system, or our interpretation of it, seems dumb, or are we ignorant?
Ok the answer to my question is yes. System locked up forcing a hard restart yesterday. Drive unrecognized on restart and Disk Utility froze trying to format it. Thanks Intel for taking an interest in this thread.
Sorry to hear.... when was the drive purchased? The warranty is 3-years from that date: http://www.intel.com/support/ssdc/hpssd/sb/CS-029645.htm http://www.intel.com/support/ssdc/hpssd/sb/CS-029645.htm