Hi all, I have recent completed my build of unRAID using the parts listed in my signature using old parts mostly and have been testing it out. I mainly use it to share data on the array and run a Win10 VM with GPU and USB pass-through. I would like a bit of help/input/advice from more knowledgeble users on the health of my drives.

 

In particular my Seagate 2 TB drive (ST2000DL003-9VT166 6YD1LNKA) has been reported as faulty twice now.

Both times on the webgui it has the read error alert in red then the orange disk is missing error. The disk will then fail to show under disk and I can't see its attribute or get SMART reports. Upon rebooting the disk will show up fine with the SMART data reporting.

First time this happen I thought I lost the SMART report and forgot to get the relevant files needed so I have only posted the files from the most recent failure. I did forget to saved the logs before again so I grabbed the saved syslog and trimmed a large block of

AMD-Vi: Event logged [iO_PAGE_FAULT device=00:13.0 domain=0x001c address=0x00000000bffdee80 flags=0x0020]

so that I could attached it. I believe the failure started from

Jan 22 13:03:39 Kerrigan kernel: ata3.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen

as the last I was on it before then, it was fine and when I came back at ~14:20 it was showing the error and did the reboot. When the drive doesn't appear this is the SMART report I get :

smartctl 6.2 2013-07-26 r3841 [x86_64-linux-4.1.13-unRAID] (local build)
Copyright (C) 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Vendor:               /2:0:0:0
Product:              
User Capacity:        600,332,565,813,390,450 bytes [600 PB]
Logical block size:   774843950 bytes
scsiModePageOffset: response length too short, resp_len=47 offset=50 bd_len=46
scsiModePageOffset: response length too short, resp_len=47 offset=50 bd_len=46
>> Terminate command early due to bad response to IEC mode page
A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.

Now I am thinking that this is more a cable/power issue as the drive was precleared with no errors before being used, although only once. But since then it has been through 2 full parity checks when I powered down the machine through the button as I could not get a feed when I was running with the single AMD GPU before. I have also done a short and extended SMART test after rebooting following both failures which both passed. After the first failure as well it rebuilt fine and passed a parity check after that as well. So in terms of doing a lot of reads and writes to the whole drive it seemed to handle it fine but it has errored out at seemingly random times after running for a while. From the SMART stats as well it didn't seem to be in bad health from what I can tell besides the crazy high age (older than manufacture date), the command timeouts and some values being close to the thresholds but haven't been reported as important for hard disk failure. The command timeout appeared after each time I opened the box up and replugged the power cables without turning off the power supply, so I do not think it is a real error? The command timeout error also used to be reported in a orange notice but has since stopped in 6.1.7 which I assume to be default behaviour.

 

Can anyone see a potential issue/clue somewhere that I am completely missing or is unable to interpret?

 

With my two other WD drives they are also pretty old in age and have some values approaching thresholds with crazy high "193 Load_Cycle_Count" values which I read are common for WD Greens. I have since used wdiddle3 to change the park time to 5mins so hopefully it will slow it down now and be fine as apparently WD said it will be fine up to 1million even though the spec sheet says 300,000. WMC1T1726117 did have multiple "200 Multi_Zone_Error_Rate" and "197 Current_Pending_Sector" but after preclearing they are all gone now. I did hot plug this drive in a clumsy manner (taking 3 tries to plugin the data and power cable) before preclearing so I am not sure if that might be what cause those issues?

Give these two have been pretty much stable so far, could I say that they don't have an problem that may lead to imminent failure soon?

 

Any feedback would be appreciated, thanks.

 

EDIT: Think I have marked this as resolved correctly, please let me know if I did it incorrectly

syslog-20160122-141918mod.txt

kerrigan-diagnostics-20160122-1432.zip

preclear_rpt_WD-WMC1T1726117_2016-01-01.txt

preclear_start_WD-WMC1T1726117_2016-01-01.txt

Can anyone see a potential issue/clue somewhere that I am completely missing or is unable to interpret?

 

SMART for the Seagate looks good, but healthy SMART does not always equal healthy disk, I would replace power/sata cables and switch sata port, if it fails again it’s probably the disk.

 

 

Give these two have been pretty much stable so far, could I say that they don't have an problem that may lead to imminent failure soon?

 

Very hard to predict, I personally don’t use any disk with previous bad sectors, as I believe they are much more likely to get more, but many do and it can be fine for years.

Can anyone see a potential issue/clue somewhere that I am completely missing or is unable to interpret?

 

SMART for the Seagate looks good, but healthy SMART does not always equal healthy disk, I would replace power/sata cables and switch sata port, if it fails again it’s probably the disk.

 

 

Give these two have been pretty much stable so far, could I say that they don't have an problem that may lead to imminent failure soon?

 

Very hard to predict, I personally don’t use any disk with previous bad sectors, as I believe they are much more likely to get more, but many do and it can be fine for years.

 

Thanks very much for your response and sorry for the massive delay in a response. Unfortunately had no time to work on my UNRAID machine for awhile.

 

So it seems like it was a cable issue in the sense of some interference from an old hot GTX275 that sat nearby squeezing the cables or just being slightly loose with that card in the way of the cables. I have now run the machine for over 3weeks with the longest run being about 10days with no issues after I removed the GTX275. I am pretty sure it is not a power issue as the machine is only drawing 200W at load.

 

With regards to the bad sectors, I see. After more reading I do gather that in general they do tend to be more unstable after 1 but at the same time have seen manufacturers not accepting a drive for failure solely due to bad sectors. I'll try to see if I can get it RMAed, but will just keep an eye on it. It did randomly obtain multiple zone errors again so yeah it is probably close to kicking it.