Hi all

 

I logged in tonight and spotted that my array flagged up a disk with 1 error on 4th Feb... its now 8th Feb and on the dashboard it shows;

I dont know whether its READ/WRITE errors - but have done a short SMART test which shows no errors logged;

 

The disk shows it has been powered up for about 3 years (25806 hours), but I would expect it to last longer.

 

I have attached the SMART report and would appreciate what your thoughts are.

 

My immediate reaction is to replace the drive and rebuild using parity?

 

Many thanks for any advice

 

WDC_WD40EFRX SMART REPORT.txt

43 minutes ago, SliMat said:

my array flagged up a disk with 1 error on 4th Feb

What exactly was the error? How did you notice it, was there a message? 

Cables / Connectors can be a culprit for random IO errors even on healthy disks

At this point, I would just run an extended smart test to be sure. If that passes, it should be ok

47 minutes ago, SliMat said:

on the dashboard it shows;

FYI, that isn't the Dashboard.

6 hours ago, apandey said:

What exactly was the error? How did you notice it, was there a message? 

Cables / Connectors can be a culprit for random IO errors even on healthy disks

At this point, I would just run an extended smart test to be sure. If that passes, it should be ok

 

Thanks for the reply @apandey

I noticed the error when I logged into the UnRAID UI - there were the usual few balloons telling me that parity check had been started, then finished and there was a yellow balloon saying that there was a disk with 1 error... then when I logged into the main screen to look at the disk status' I spotted disk #3 showing 34 errors. the popup balloon didnt give any more information than that.

 

I have just started an extended SMART test now and will check the results later.

 

6 hours ago, trurl said:

FYI, that isn't the Dashboard.

Sorry @trurl - poor terminology... I meant 'main UI'

 

4 hours ago, SliMat said:

there were the usual few balloons telling me that parity check had been started, then finished and there was a yellow balloon saying that there was a disk with 1 error

Usually it says what type of error. Was there a corresponding green notification saying disk error returned to normal? If it happens again, take note. For now, let's see the result of extended test

If you haven't rebooted post diagnostics. They might tell more than just the SMART report. Connection problems, for example, are more likely to be seen in syslog than SMART report.

 

And in general, we prefer complete diagnostics instead of just SMART reports or syslog. Diagnostics includes all these and more.

1 hour ago, trurl said:

If you haven't rebooted post diagnostics

 

No, I haven't rebooted and the main screen still shows 34 errors... diags are attached.

 

The extended SMART test is still running - showing at 90% complete... so when its finished I will post this too.

 

Thanks

unraid-diagnostics-20230208-1424.zip

Probably unrelated but your syslog is being spammed with some things my text editor can't even display accurately

Jan  5 11:08:35 UnRAID smbd[34949]: [2023/01/05 11:08:35.539553,  0] ../../lib/param/loadparm.c:1841(lpcfg_do_global_parameter)
Jan  5 11:08:35 UnRAID smbd[34949]:   Ignoring unknown parameter "veto files"

Seems to be related to your smb-extra.conf. Maybe a corrupt file on flash?

# Stop the annoying ._ / ._DS_Store files being written when Mac OS accesses UnRAID shares
ve?to files = /._*/.DS_St?ore/?

You might check SMB Extras in SMB Settings.

1 hour ago, trurl said:

Probably unrelated but your syslog is being spammed with some things my text editor can't even display accurately

 

Odd, it looks like some sort of corruption in syslog1.txt.... if you look in syslog.txt it reads perfectly;

Jan 18 08:50:11 UnRAID emhttpd: read SMART /dev/sdg
Jan 18 08:50:29 UnRAID smbd[30230]: [2023/01/18 08:50:29.331627,  0] ../../lib/param/loadparm.c:801(lpcfg_map_parameter)
Jan 18 08:50:29 UnRAID smbd[30230]:   Unknown parameter encountered: "veto files"
Jan 18 08:50:29 UnRAID smbd[30230]: [2023/01/18 08:50:29.331672,  0] ../../lib/param/loadparm.c:1841(lpcfg_do_global_parameter)
Jan 18 08:50:29 UnRAID smbd[30230]:   Ignoring unknown parameter "veto files"

 

This was something to stop the SMB share filling up with .DS_Store files everytime a Mac accessed files on UnRAID...

https://forums.unraid.net/topic/48257-how-to-stop-ds_store-files-and-_-files-being-created-in-shares-server-side/

 

The smb-extra.conf file reads perfectly in Notepad++;

# Stop the annoying ._ / ._DS_Store files being written when Mac OS accesses UnRAID shares
veto files = /._*/.DS_Store/

 

To be honest it was so long ago that I set this up, I cant remember the tutorial I followed  So, will it hurt to empty (or delete) smb-extra.conf from  \\xxx.xxx.xxx.xxx\flash\config\ as I have no Macs on the system now?

 

Edited February 8, 20233 yr by SliMat

Try clearing it from SMB Extras. Might have to stop/start array. Go ahead and let extended self-test complete.

OK, the extended self test is complete and says finished with no errors - see attached.

 

So, you think I should just remove the contents from smb-extra.conf and restart the array?

 

When I saw the errors on the HDD I was worried it was about to fail and run the risk of data loss

 

20 hours ago, apandey said:

Cables / Connectors can be a culprit for random IO errors even on healthy disks

I forgot to say that the machine is an HPE Proliant ML320e v2 which is using a P410i control card in HBA mode - this connects to the HDD cage using mini-SAS cables, so I am pretty confident its not a cable / connector issue.

 

Thanks for all the help and advice guys

unraid-smart-20230208-2225.zip

59 minutes ago, SliMat said:

extended self test is complete and says finished with no errors - see attached.

Looks fine

 

1 hour ago, SliMat said:

you think I should just remove the contents from smb-extra.conf and restart the array?

yes

 

errors start here. Not sure if connection problem or not. Since it passed extended test should be OK, and it wasn't disabled. Main - Array Operation - Clear Stats.

Feb  4 20:38:03 UnRAID kernel: ata4.00: exception Emask 0x0 SAct 0x80c001ff SErr 0x0 action 0x0
Feb  4 20:38:03 UnRAID kernel: ata4.00: irq_stat 0x40000008
Feb  4 20:38:03 UnRAID kernel: ata4.00: failed command: READ FPDMA QUEUED
Feb  4 20:38:03 UnRAID kernel: ata4.00: cmd 60/10:b0:98:f8:a1/01:00:6c:01:00/40 tag 22 ncq dma 139264 in
Feb  4 20:38:03 UnRAID kernel:         res 41/40:00:e0:f8:a1/00:00:6c:01:00/40 Emask 0x409 (media error) <F>
Feb  4 20:38:03 UnRAID kernel: ata4.00: status: { DRDY ERR }
Feb  4 20:38:03 UnRAID kernel: ata4.00: error: { UNC }
Feb  4 20:38:03 UnRAID kernel: ata4.00: configured for UDMA/133
Feb  4 20:38:03 UnRAID kernel: sd 7:0:0:0: [sdh] tag#22 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08 cmd_age=0s
Feb  4 20:38:03 UnRAID kernel: sd 7:0:0:0: [sdh] tag#22 Sense Key : 0x3 [current] 
Feb  4 20:38:03 UnRAID kernel: sd 7:0:0:0: [sdh] tag#22 ASC=0x11 ASCQ=0x4 
Feb  4 20:38:03 UnRAID kernel: sd 7:0:0:0: [sdh] tag#22 CDB: opcode=0x88 88 00 00 00 00 01 6c a1 f8 98 00 00 01 10 00 00
Feb  4 20:38:03 UnRAID kernel: blk_update_request: I/O error, dev sdh, sector 6117521560 op 0x0:(READ) flags 0x0 phys_seg 34 prio class 0

 

OK, rebooted after commenting out contents of smb-extra.conf and got this;

 

Odd that after so many years smb-extra.conf could cause this, so I am not 100% sure that this was the fault but at least I am reassured that disk 3 looks OK

 

Many thanks all.