May 11, 20188 yr Hello - oh wise unraid peoples - Does anyone know of an app or program that interprets SMART files? I hate to bother people here continuously.
May 11, 20188 yr 6 hours ago, cbr600ds2 said: Hello - oh wise unraid peoples - Does anyone know of an app or program that interprets SMART files? I hate to bother people here continuously. On one hand "all" programs interprets SMART data. I.e. they translate the SMART data into information. On the other hand "no" programs interprets SMART data. I.e. unless the drive itself explicitly says "fail", they mostly leave it to the user to decide where to put the limit between "good enough" and "replace". One of the reasons for this is that not even the disk manufacturers are fully agreeing on the SMART data. They may give score of "perfect" for a specific attribute even when the drive is almost dead - just because they prefer to only do warranty replacements of truly dead drives. So they more or less intentionally let people lose data just to optimize their return process. For the most part, I think you are best off posting your SMART data here - only then can you hope to not just get feedback on the values, but also get feedback on why people think a value is dangerous or not. This is also a great way of learning - when listening to the suggestion of "magic software", everything will always be black magic.
May 11, 20188 yr For a quick lesson in SMART data evaluation. First go to Settings >>>> Disk Settings, now stroll down to the section entitled "GLOBAL SMART SETTINGS". You will see a list of six SMART attributes. If any of them are not zero, you should be concerned. (By the way, you should have set up Notifications to notify you whenever one of these occurs!!!) WikiPedia has an excellent article on SMART that you can find here: https://en.wikipedia.org/wiki/S.M.A.R.T. Read about each of those attributes. One thing, you will find is that not all hard drives even report all of these six parameters! Attribute 199 (UDMA CRC Error Count) is not actually a disk error at all--- 90+% of all attribute199 errors are cabling problems! Attribute 5 (Reallocated Sectors Count) is not a true failure but if it is increasing, it is an good indicator that the disk is on its way to a failure. (Many of us are paranoid enough about this one that we will retire any disk to some other duty when the first one pops up.) As @pwm pointed out, the SMART data is intended to minimize warranty returns. for example, they assume that a few dozen Reallocated Sectors won't affect most people as they could be on a region of the disk where there isn't any data stored. But, we unRAID users, have to be able to read EVERY mapped sector on a disk to do any parity type operation.
May 16, 20188 yr Author Hm.... I only ask because i have one disk that I keep having issues with because it'll fail but then pass Smart test. None of the attributes I have are set to 0. Should I change it? I've actually been thinking that it could be an issue w/ just the one part of the break out cable since it fails and then passes so often. Should I post the info?
May 16, 20188 yr Author Maybe its time to upgrade breakout cables... they must make 6 GB breakout cables right? WL2000GSA6472C_WOL240266532-20180515-2136.txt
May 16, 20188 yr 4 hours ago, cbr600ds2 said: Maybe its time to upgrade breakout cables... they must make 6 GB breakout cables right? WL2000GSA6472C_WOL240266532-20180515-2136.txt On one hand it shows good SMART counter values. But you have never done a long SMART test - the conveyance test was done when the drive was brand new and is just a quick test to check for transport damage. Only the extended test will read through every sector on the drive. The SMART data doesn't show any UDMA CRC errors which are normally an indication of transfer errors often caused by the cable. Anyway - you complain about having constant problems with the drive but don't mention what problems and you haven't posted any logs. So we don't know if your issues might be that it's maybe connected to a Marvell controller that now and then disconnects the drive because of issues with the controller and not issues with the drive. You do have this in the SMART data, that indicates a communications disconnect: 0x0009 2 3 Transition from drive PhyRdy to drive PhyNRdy 0x000a 2 1 Device-to-host register FISes sent due to a COMRESET
Archived
This topic is now archived and is closed to further replies.