SMART error display logic

Hey guys, I'm new to OMV, been using it for about a month. Originally tested it on the Raspberry Pi and loved it so much that I decided to build a NAS box with spare parts I had laying around.

I did invest into hard drives and went for a cheap option I am now regretting, these were 2TB hitachi drives that were going for 25 bucks. My plan was to do a RAID 5 with them but just hit a wall for which I decided to open this thread.

What I am trying to find out is the logic used for reporting SMART errors in the interface and their actual impact on the overall setup. Allow me to elaborate:

One of the drives I ordered has 4 smart errors. Debugging the smart information I realized there was 1 block pending allocation so I decided to do a full format a force that block to be written to so it would rellocate. After doing that I ran another SMART long test and sure enough the pending allocation error count was now gone, however the OMV GUI is still showing up red with smart errors.

I am pasting my current SMART output for reference bellow (I am right now in the middle of running another long test). What I am trying to figure out is what sort of logic is used to determine if the drive is usable or not since it does appear to PASS the test successfully (without errors). Does this means once there is a smart error it wont go back to green even after the drive self fixes? I suppose that my worry here is that since the red light is not going away it makes me somewhat hesitant to rely on the drive but the fact that the lon SMART test passes successfully sort of contradicts the red light, so I am trying to make an informed decision on how I should proceed.

It would be a lot simpler to just return the drive and move on (and I am likely to eventually do that) but I also feel this is a learning experience that I can't pass away. Have already learned quite a bit I didn't know on how SMART works and how to read the results it yields (somewhat).

Thanks!

NOTE: the board complained about the thread being too long so I am attaching smart.txt.

Zitat von ryecoaaron

Read this. It explains the smart statistics to watch. The first thing I noticed was your relocated sector count was very high. That is one of the backblaze statistics to watch as well. I wouldn't trust those drives especially in a raid array.

Fantastic article, thanks so much for sharing it.

When I saw this in the comments I literally facepalm when I rechecked my report:

Zitat

SMART 5: Reallocated_Sector_Count
1-4 keep an eye on it, more than 4 replace

This drive is certainly not reliable. The other two have no records so I am hoping they are okay. So what I'll do is return the bad one and order a new one so I can do the RAID5 and I'll use the other two as stand alone in the meantime to shuffle around some data I have to compile from other drives.

@ryecoaaron Just so I can make peace with the topic at hand, if the Reallocated_Sector_Count was low, would that have played a role in getting a green light on the SMART monitoring tool in the GUI? or would the tool always report a red light forever once an error is found (even if fixed?)