just recently updated to the v6. was in the middle of doing a parity check which was running at less 1MBs when the check stopped and one of my disks was flagged as unmountable. replaced the disk, rebuilt the data, corrected the file system issues. ran a new parity check and am still getting speeds of less than 1MBs.

 

Diagnostics log is too large to attach. link to dropbox to dl.

 

https://www.dropbox.com/sh/efm0bymjjnz9wtr/AADoNHWOeaZ1HExMrA1VZRYza?dl=0

Hi -

 

What are the specs on your current system?  unRAID 6 has slow parity check speeds on some very low end systems that worked Ok with unRAID 5.

Syslog is full of errors for you parity disk, like these:

Apr 26 09:59:32 tower kernel: ata12.00: status: { Busy }
Apr 26 09:59:32 tower kernel: ata12.00: error: { UNC AMNF IDNF ABRT }
Apr 26 09:59:32 tower kernel: ata12: hard resetting link
Apr 26 09:59:38 tower kernel: ata12: SATA link up 1.5 Gbps (SStatus 113 SControl 10)
Apr 26 09:59:38 tower kernel: ata12.00: configured for UDMA/33
Apr 26 09:59:38 tower kernel: ata12: EH complete
Apr 26 09:59:38 tower kernel: ata12.00: exception Emask 0x10 SAct 0x0 SErr 0x80000 action 0xe frozen
Apr 26 09:59:38 tower kernel: ata12.00: irq_stat 0x00100010, PHY RDY changed
Apr 26 09:59:38 tower kernel: ata12: SError: { 10B8B }
Apr 26 09:59:38 tower kernel: ata12.00: failed command: READ DMA EXT
Apr 26 09:59:38 tower kernel: ata12.00: cmd 25/00:e8:c8:97:b6/00:07:02:00:00/e0 tag 0 dma 1036288 in
Apr 26 09:59:38 tower kernel:         res ee/75:00:00:00:00/00:00:80:0f:ee/00 Emask 0x12 (ATA bus error)
Apr 26 09:59:38 tower kernel: ata12.00: status: { Busy }
Apr 26 09:59:38 tower kernel: ata12.00: error: { UNC AMNF IDNF ABRT }
Apr 26 09:59:38 tower kernel: ata12: hard resetting link
Apr 26 09:59:44 tower kernel: ata12: SATA link up 1.5 Gbps (SStatus 113 SControl 10)
Apr 26 09:59:44 tower kernel: ata12.00: configured for UDMA/33
Apr 26 09:59:44 tower kernel: ata12: EH complete
Apr 26 09:59:44 tower kernel: ata12.00: exception Emask 0x10 SAct 0x0 SErr 0x80000 action 0xe frozen
Apr 26 09:59:44 tower kernel: ata12.00: irq_stat 0x00100010, PHY RDY changed
Apr 26 09:59:44 tower kernel: ata12: SError: { 10B8B }
Apr 26 09:59:44 tower kernel: ata12.00: failed command: READ DMA EXT
Apr 26 09:59:44 tower kernel: ata12.00: cmd 25/00:80:10:c2:b6/00:07:02:00:00/e0 tag 0 dma 983040 in
Apr 26 09:59:44 tower kernel:         res ee/75:00:00:00:00/00:00:80:0d:ee/00 Emask 0x12 (ATA bus error)
Apr 26 09:59:44 tower kernel: ata12.00: status: { Busy }
Apr 26 09:59:44 tower kernel: ata12.00: error: { UNC AMNF IDNF ABRT }
Apr 26 09:59:44 tower kernel: ata12: hard resetting link
Apr 26 09:59:49 tower kernel: ata12: SATA link up 1.5 Gbps (SStatus 113 SControl 10)
Apr 26 09:59:49 tower kernel: ata12.00: configured for UDMA/33
Apr 26 09:59:49 tower kernel: ata12: EH complete
Apr 26 09:59:50 tower kernel: ata12.00: exception Emask 0x10 SAct 0x0 SErr 0x80000 action 0xe frozen
Apr 26 09:59:50 tower kernel: ata12.00: irq_stat 0x00100010, PHY RDY changed
Apr 26 09:59:50 tower kernel: ata12: SError: { 10B8B }
Apr 26 09:59:50 tower kernel: ata12.00: failed command: READ DMA EXT
Apr 26 09:59:50 tower kernel: ata12.00: cmd 25/00:e8:f0:5d:b7/00:07:02:00:00/e0 tag 0 dma 1036288 in
Apr 26 09:59:50 tower kernel:         res ee/75:00:00:00:00/00:00:c0:0d:ee/00 Emask 0x12 (ATA bus error)
Apr 26 09:59:50 tower kernel: ata12.00: status: { Busy }
Apr 26 09:59:50 tower kernel: ata12.00: error: { UNC AMNF IDNF ABRT }
Apr 26 09:59:50 tower kernel: ata12: hard resetting link

 

SMART for it looks ok, I'd start by trying different controller/cable.

Not trying to hijack this thread but read it because I have a similar problem. I upgraded to 6 the used unbalance to move files to reformat to XFS and monthly parity check went okay I believe after the conversion. I recently also upgraded this server to match my other server with a new and same motherboard processor and memory. Other identical server runs great. This one has problems and now wont finish parity. It is very slow at times and very fast at others. The slowness to me must mean errors because there have been lots. But not sure why. As I write this I'm thinking cable(s) as mentioned above. Could it be? Yeah but you folks can read the logs much better then I can make them out some but sure could use your help. If it's the processor or MB that you find at fault I have only a couple of days to swap it out. I also have done a 20 hour memory test without an error and ran it 4 round based on the display readout.

Any help you have is much appreciated thank you,

and any more help I can be please let me know

Full System log file to large to upload:

Drop box link below.

https://www.dropbox.com/s/i44kbrizydnds1l/r2d2-diagnostics-20160426-2034.zip?dl=0

 

 

MB- MSIZ97 Gaming 5

Processor- Intel Devils Canyon 4790K

Memeory- 32G

Sata Card is 8 out SuperMicro

Not trying to hijack this thread but read it because I have a similar problem. I upgraded to 6 the used unbalance to move files to reformat to XFS and monthly parity check went okay I believe after the conversion. I recently also upgraded this server to match my other server with a new and same motherboard processor and memory. Other identical server runs great. This one has problems and now wont finish parity. It is very slow at times and very fast at others. The slowness to me must mean errors because there have been lots. But not sure why. As I write this I'm thinking cable(s) as mentioned above. Could it be? Yeah but you folks can read the logs much better then I can make them out some but sure could use your help. If it's the processor or MB that you find at fault I have only a couple of days to swap it out. I also have done a 20 hour memory test without an error and ran it 4 round based on the display readout.

Any help you have is much appreciated thank you,

and any more help I can be please let me know

Full System log file to large to upload:

Drop box link below.

https://www.dropbox.com/s/i44kbrizydnds1l/r2d2-diagnostics-20160426-2034.zip?dl=0

 

 

MB- MSIZ97 Gaming 5

Processor- Intel Devils Canyon 4790K

Memeory- 32G

Sata Card is 8 out SuperMicro

First guess is the SAS2LP, try with virtualization disable if possible or another controller if available.

Not trying to hijack this thread but read it because I have a similar problem. I upgraded to 6 the used unbalance to move files to reformat to XFS and monthly parity check went okay I believe after the conversion. I recently also upgraded this server to match my other server with a new and same motherboard processor and memory. Other identical server runs great. This one has problems and now wont finish parity. It is very slow at times and very fast at others. The slowness to me must mean errors because there have been lots. But not sure why. As I write this I'm thinking cable(s) as mentioned above. Could it be? Yeah but you folks can read the logs much better then I can make them out some but sure could use your help. If it's the processor or MB that you find at fault I have only a couple of days to swap it out. I also have done a 20 hour memory test without an error and ran it 4 round based on the display readout.

Any help you have is much appreciated thank you,

and any more help I can be please let me know

Full System log file to large to upload:

Drop box link below.

https://www.dropbox.com/s/i44kbrizydnds1l/r2d2-diagnostics-20160426-2034.zip?dl=0

 

 

MB- MSIZ97 Gaming 5

Processor- Intel Devils Canyon 4790K

Memeory- 32G

Sata Card is 8 out SuperMicro

First guess is the SAS2LP, try with virtualization disable if possible or another controller if available.

 

I have also tried without plugin's enabled and don't have any visualization running that I know of or dockers on. I'm swapping out the card and cables and trying again. Time will tell and will report back.

I other thing I forgot to mention is each time I run parity the server looses network and also does freeze up the system. And I have to do a force shutdown.

Thank you for your help,

 

just want to report back. swapped my parity to my supermicro controller card and now getting steady 60mbs rates. i think my old sata card is going bad. prob time to replace it with something new.

Not trying to hijack this thread but read it because I have a similar problem. I upgraded to 6 the used unbalance to move files to reformat to XFS and monthly parity check went okay I believe after the conversion. I recently also upgraded this server to match my other server with a new and same motherboard processor and memory. Other identical server runs great. This one has problems and now wont finish parity. It is very slow at times and very fast at others. The slowness to me must mean errors because there have been lots. But not sure why. As I write this I'm thinking cable(s) as mentioned above. Could it be? Yeah but you folks can read the logs much better then I can make them out some but sure could use your help. If it's the processor or MB that you find at fault I have only a couple of days to swap it out. I also have done a 20 hour memory test without an error and ran it 4 round based on the display readout.

Any help you have is much appreciated thank you,

and any more help I can be please let me know

Full System log file to large to upload:

Drop box link below.

https://www.dropbox.com/s/i44kbrizydnds1l/r2d2-diagnostics-20160426-2034.zip?dl=0

 

 

MB- MSIZ97 Gaming 5

Processor- Intel Devils Canyon 4790K

Memeory- 32G

Sata Card is 8 out SuperMicro

First guess is the SAS2LP, try with virtualization disable if possible or another controller if available.

 

I have also tried without plugin's enabled and don't have any visualization running that I know of or dockers on. I'm swapping out the card and cables and trying again. Time will tell and will report back.

I other thing I forgot to mention is each time I run parity the server looses network and also does freeze up the system. And I have to do a force shutdown.

Thank you for your help,

Then check if virtualization in on in the board bios, if yes disable, usually it's called  VT-D.

I run 6 SAS2LP cards in 2 servers and have no parity speed issues with v6. You do need to increase the Tunable (md_num_stripes) and Tunable (md_sync_window) settings on many setups to get the same speeds as v5. I had to use 5632 and 2560 respectively to return to v5 speeds.

 

No idea if virtualization is disabled in BIOS, I assume not because i'm currently running a Win10 VM on the same system.

I run 6 SAS2LP cards in 2 servers and have no parity speed issues with v6. You do need to increase the Tunable (md_num_stripes) and Tunable (md_sync_window) settings on many setups to get the same speeds as v5. I had to use 5632 and 2560 respectively to return to v5 speeds.

 

No idea if virtualization is disabled in BIOS, I assume not because i'm currently running a Win10 VM on the same system.

The SAS2LP issue only affects a small number of users, most, including myself, have no issues with it, and the main issue are DMA errors, that besides slowing  down the parity check can drop disks in some cases.

 

Turning vt-d off solves the issue for some.

just want to report back. swapped my parity to my supermicro controller card and now getting steady 60mbs rates. i think my old sata card is going bad. prob time to replace it with something new.

It might be going bad - but maybe not.  I also had stability problems when I migrated to unRAID 6 and wound up moving parity onto my SAS2LP as one of the steps I took to stabilize things (one of several steps).  That resulted in slower parity check speeds (but not as slow as yours) until recently when Limetech introduced a fix.  The SAS2LP is generally working better for most folks now, but I still think there are some compatibility issues on older hardware like mine.

Thank you all for the help I found the problem to be the 2 SAS Sata Cables or at least one is bad. Nice thing about them is the lifetime warranty. But purchased new ones to find out that they are the problem. Parity check is back at its top speeds and finishing correctly. After that also upgraded parity to 6TB to that server and completed fast and without issue. System has been on and stable for the last week now steaming from and running dockers and addon's with fast results.

Fingers crossed smooth sailing for all...