SAS Disk Error - Diagnosis Help Please

[Knights21]

Disc 3 in my array came up with a lot of read errors so I swapped out the disc and rebuilt the array only for the replacement disc to report read errors a few days later. Although the replacement was not a new disc, made me suspicious that it was not actually the disc. Would someone far cleverer than me take a look at the two smart logs for me and determine if it is the discs or a power issue? 😎

Earlier date smart is previous disc.

TIA.

 

HP ProLiant Gen 8

tower-smart-20220807-1021.zip tower-smart-20221028-1640.zip tower-diagnostics-20221028-1700.zip

[JorgeB]

Syslog is completely spammed with smb related info, but SMART for both disks looks OK, so likely not a disk problem, check/replace cables or swap slots.

[bing281]

you can get errors on a disk that is considered out of parity if another disk is losing sectors basically you need to look at the SMART attributes on all the drives you have in the array including the parity drive and see if one of them is getting more REALLOCATED SECTORS.   A lost sector on any drive during parity rebuild will fail the rebuild 

[Knights21]

None of the other discs report errors in the main window of the UI. 

Will try a bay swap. Thanks 👍

[bing281]

Your problem is with this disk 

 

=== START OF INFORMATION SECTION ===
Vendor:               HITACHI
Product:              HUS723030ALS640
Revision:             A440
Compliance:           SPC-4
User Capacity:        3,000,592,982,016 bytes [3.00 TB]
Logical block size:   512 bytes
Rotation Rate:        7200 rpm
Form Factor:          3.5 inches
Logical Unit id:      0x5000cca03ec60450
Serial number:        YVKHWZ3K
Device type:          disk
Transport protocol:   SAS (SPL-4)
Local Time is:        Fri Oct 28 17:01:36 2022 BST
SMART support is:     Available - device has SMART capability.
SMART support is:     Enabled
Temperature Warning:  Enabled
Read Cache is:        Enabled
Writeback Cache is:   Enabled

 

and the problem is here 

 

Protocol Specific port log page for SAS SSP
relative target port id = 1
  generation code = 3
  number of phys = 1
  phy identifier = 0
    attached device type: SAS or SATA device
    attached reason: unknown
    reason: unknown
    negotiated logical link rate: phy enabled; 6 Gbps
    attached initiator port: ssp=1 stp=1 smp=1
    attached target port: ssp=0 stp=0 smp=0
    SAS address = 0x5000cca03ec60451
    attached SAS address = 0x51402ec0002d2a34
    attached phy identifier = 4
    Invalid DWORD count = 1327
    Running disparity error count = 1300
    Loss of DWORD synchronization count = 35
    Phy reset problem count = 0
relative target port id = 2
  generation code = 3
  number of phys = 1
  phy identifier = 1
    attached device type: no device attached
    attached reason: unknown
    reason: power on
    negotiated logical link rate: phy enabled; unknown
    attached initiator port: ssp=0 stp=0 smp=0
    attached target port: ssp=0 stp=0 smp=0
    SAS address = 0x5000cca03ec60452
    attached SAS address = 0x0
    attached phy identifier = 0
    Invalid DWORD count = 0
    Running disparity error count = 0
    Loss of DWORD synchronization count = 0
    Phy reset problem count = 0

 

This device doesn't give you back smart attributes because of the error above which looks like it is losing connection to the SAS card.   

[bing281]

actually it looks like disks 1,2,3 are all showing this issue.   Are those 3 disks connected differently than the others.   Looking at the LSPCI log it looks like they are connected to the system via USB? 

[Knights21]

The three SAS drives are connected via a HP HBA card whereas the parity and cache via the onboard SATA controller.

[Knights21]

On 10/28/2022 at 5:40 PM, bing281 said:

This device doesn't give you back smart attributes because of the error above which looks like it is losing connection to the SAS card.   

Perhaps I should revert back to SATA drives and ditch the HBA?

[Knights21]

After the latest UnRaid update the problem seems to have resolved itself. Still using the second drive but now no read errors at all. Two parity checks done since the update.