Hi all!

 

Just got my unRaid server setup and I'm having some hiccups when browsing the webGui and my Plex library.  They tend to become unresponsive for a couple of moments (sometimes 30secs, sometimes 2mins).  Upon checking my system log and my hard drive logs, I found this on my cache drives. 

 

May 31 12:44:12 Tower kernel: ata8: SATA max UDMA/133 abar m2048@0xdf910000 port 0xdf910180 irq 46
May 31 12:44:12 Tower kernel: ata8: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
May 31 12:44:12 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 12:44:12 Tower kernel: ata8.00: READ LOG DMA EXT failed, trying unqueued
May 31 12:44:12 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 12:44:12 Tower kernel: ata8.00: ATA-9: Samsung SSD 850 EVO 500GB, S21HNXAG303357A, EMT01B6Q, max UDMA/133
May 31 12:44:12 Tower kernel: ata8.00: 976773168 sectors, multi 1: LBA48 NCQ (depth 31/32), AA
May 31 12:44:12 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 12:44:12 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 12:44:12 Tower kernel: ata8.00: configured for UDMA/133
May 31 12:44:12 Tower kernel: sd 9:0:0:0: [sdh] 976773168 512-byte logical blocks: (500 GB/466 GiB)
May 31 12:44:12 Tower kernel: sd 9:0:0:0: [sdh] 4096-byte physical blocks
May 31 12:44:12 Tower kernel: sd 9:0:0:0: [sdh] Write Protect is off
May 31 12:44:12 Tower kernel: sd 9:0:0:0: [sdh] Mode Sense: 00 3a 00 00
May 31 12:44:12 Tower kernel: sd 9:0:0:0: [sdh] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
May 31 12:44:12 Tower kernel: sdh: sdh1
May 31 12:44:12 Tower kernel: sd 9:0:0:0: [sdh] Attached SCSI disk
May 31 12:44:49 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:44:49 Tower emhttp: import 24 cache device: sdh
May 31 12:44:49 Tower emhttp: shcmd (7): /usr/sbin/hdparm -S0 /dev/sdh &> /dev/null
May 31 12:44:49 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:44:49 Tower emhttp: import 24 cache device: sdh
May 31 12:45:17 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:45:17 Tower emhttp: import 24 cache device: sdh
May 31 12:45:20 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:45:20 Tower emhttp: import 24 cache device: sdh
May 31 12:45:22 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:45:22 Tower emhttp: import 24 cache device: sdh
May 31 12:45:49 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:45:49 Tower emhttp: import 24 cache device: sdh
May 31 12:46:19 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:46:19 Tower emhttp: import 24 cache device: sdh
May 31 12:46:19 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:46:19 Tower emhttp: import 24 cache device: sdh
May 31 12:46:35 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:46:35 Tower emhttp: import 24 cache device: sdh
May 31 12:47:22 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:47:22 Tower emhttp: import 24 cache device: sdh
May 31 12:47:35 Tower emhttp: Samsung_SSD_850_EVO_500GB_S21HNXAG303357A (sdh) 488386584
May 31 12:47:35 Tower emhttp: import 24 cache device: sdh
May 31 12:47:35 Tower emhttp: shcmd (61): /usr/sbin/hdparm -S0 /dev/sdh &> /dev/null
May 31 12:47:35 Tower kernel: BTRFS: device fsid a64386f3-4ec6-4195-8fb8-30a743718631 devid 2 transid 6837 /dev/sdh1
May 31 13:06:06 Tower kernel: ata8.00: exception Emask 0x0 SAct 0x1000 SErr 0x0 action 0x6 frozen
May 31 13:06:06 Tower kernel: ata8.00: failed command: WRITE FPDMA QUEUED
May 31 13:06:06 Tower kernel: ata8.00: cmd 61/10:60:78:69:f2/00:00:10:00:00/40 tag 12 ncq 8192 out
May 31 13:06:06 Tower kernel: ata8.00: status: { DRDY }
May 31 13:06:06 Tower kernel: ata8: hard resetting link
May 31 13:06:06 Tower kernel: ata8: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
May 31 13:06:06 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 13:06:06 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 13:06:06 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 13:06:06 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 13:06:06 Tower kernel: ata8.00: configured for UDMA/133
May 31 13:06:06 Tower kernel: ata8.00: device reported invalid CHS sector 0
May 31 13:06:06 Tower kernel: ata8: EH complete
May 31 13:06:38 Tower kernel: ata8.00: exception Emask 0x0 SAct 0x20000 SErr 0x0 action 0x6 frozen
May 31 13:06:38 Tower kernel: ata8.00: failed command: WRITE FPDMA QUEUED
May 31 13:06:38 Tower kernel: ata8.00: cmd 61/10:88:08:55:f5/00:00:10:00:00/40 tag 17 ncq 8192 out
May 31 13:06:38 Tower kernel: ata8.00: status: { DRDY }
May 31 13:06:38 Tower kernel: ata8: hard resetting link
May 31 13:06:38 Tower kernel: ata8: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
May 31 13:06:38 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 13:06:38 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 13:06:38 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 13:06:38 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 13:06:38 Tower kernel: ata8.00: configured for UDMA/133
May 31 13:06:38 Tower kernel: ata8.00: device reported invalid CHS sector 0
May 31 13:06:38 Tower kernel: ata8: EH complete
May 31 13:33:36 Tower kernel: ata8.00: exception Emask 0x0 SAct 0xc800000 SErr 0x0 action 0x6 frozen
May 31 13:33:36 Tower kernel: ata8.00: failed command: READ FPDMA QUEUED
May 31 13:33:36 Tower kernel: ata8.00: cmd 60/08:b8:a0:bb:a4/00:00:08:00:00/40 tag 23 ncq 4096 in
May 31 13:33:36 Tower kernel: ata8.00: status: { DRDY }
May 31 13:33:36 Tower kernel: ata8.00: failed command: READ FPDMA QUEUED
May 31 13:33:36 Tower kernel: ata8.00: cmd 60/08:d0:70:cc:d4/00:00:08:00:00/40 tag 26 ncq 4096 in
May 31 13:33:36 Tower kernel: ata8.00: status: { DRDY }
May 31 13:33:36 Tower kernel: ata8.00: failed command: READ FPDMA QUEUED
May 31 13:33:36 Tower kernel: ata8.00: cmd 60/08:d8:20:ce:09/00:00:09:00:00/40 tag 27 ncq 4096 in
May 31 13:33:36 Tower kernel: ata8.00: status: { DRDY }
May 31 13:33:36 Tower kernel: ata8: hard resetting link
May 31 13:33:36 Tower kernel: ata8: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
May 31 13:33:36 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 13:33:36 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 13:33:36 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 13:33:36 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 13:33:36 Tower kernel: ata8.00: configured for UDMA/133
May 31 13:33:36 Tower kernel: ata8.00: device reported invalid CHS sector 0
May 31 13:33:36 Tower kernel: ata8.00: device reported invalid CHS sector 0
May 31 13:33:36 Tower kernel: ata8.00: device reported invalid CHS sector 0
May 31 13:33:36 Tower kernel: ata8: EH complete
May 31 17:41:57 Tower kernel: ata8.00: exception Emask 0x0 SAct 0x4000000 SErr 0x0 action 0x6 frozen
May 31 17:41:57 Tower kernel: ata8.00: failed command: WRITE FPDMA QUEUED
May 31 17:41:57 Tower kernel: ata8.00: cmd 61/10:d0:78:90:65/00:00:0b:00:00/40 tag 26 ncq 8192 out
May 31 17:41:57 Tower kernel: ata8.00: status: { DRDY }
May 31 17:41:57 Tower kernel: ata8: hard resetting link
May 31 17:41:57 Tower kernel: ata8: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
May 31 17:41:57 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 17:41:57 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 17:41:57 Tower kernel: ata8.00: supports DRM functions and may not be fully accessible
May 31 17:41:57 Tower kernel: ata8.00: failed to get NCQ Send/Recv Log Emask 0x1
May 31 17:41:57 Tower kernel: ata8.00: configured for UDMA/133
May 31 17:41:57 Tower kernel: ata8.00: device reported invalid CHS sector 0
May 31 17:41:57 Tower kernel: ata8: EH complete

 

Upon some Googling around, I found that this error could be caused by a bad drive or bad cables.  Can anyone tell me by the logs at which one I should be going after first? Or am I barking up the wrong tree?  The drives in question are Samsung EVO 850 500gb SSDs. Thanks for any light you can share on this issue!

Read here and post your diagnostics.

Diagnostic logs attached

tower-diagnostics-20160531-1810.zip

Bump

Been Googling around and I think I found the answer.  Seems this is a pretty common problem on Linux w/ SSDs.  So, I'm posting here in case someone else goes searching the forums for the answer.  Change this setting: Settings -> Disk Settings -> Force NCQ Disabled to yes or alternatively, nano /boot/syslinux/syslinux.cfg and add libata.force=noncq to your append line.  Set the setting either way and reboot.  Will post back in a day or so if this permanently fixed my problem w/ my cache SSDs

So, I had about 10 Plex users on last night and a couple on today and that error has disappeared from my logs and no complaints from the users.  Seems that was the fix.