October 18, 20169 yr Author That conf file did force Tower to be Master, but it didn't solve the problem. Still getting the same failed file transfers to and from the Tower On the Windows computer, go to 'Control Panel', then 'Credential Manager' and see if you have any credentials set. (However, if 6.1 was working, I find it hard to believe that this is the issue but it is quick and easy to check.) Do you have any other computers that you could use to help isolate the problem? Have you changed the 'SMB Security Settings' from the defaults? The reason I am puzzled is that most of these types of issues have been occurring with the Windows 10 OS. Windows 7 has been pretty much trouble free... To add insult to injury, while testing back and forth between builds, the USB drive died. All USB fixer tools have reported it unrepairable. Now I need to wait a day for my new USB and then I'll try a fresh built of 6.2.1. We'll see how that goes...
October 18, 20169 yr Community Expert That conf file did force Tower to be Master, but it didn't solve the problem. Still getting the same failed file transfers to and from the Tower On the Windows computer, go to 'Control Panel', then 'Credential Manager' and see if you have any credentials set. (However, if 6.1 was working, I find it hard to believe that this is the issue but it is quick and easy to check.) Do you have any other computers that you could use to help isolate the problem? Have you changed the 'SMB Security Settings' from the defaults? The reason I am puzzled is that most of these types of issues have been occurring with the Windows 10 OS. Windows 7 has been pretty much trouble free... To add insult to injury, while testing back and forth between builds, the USB drive died. All USB fixer tools have reported it unrepairable. Now I need to wait a day for my new USB and then I'll try a fresh built of 6.2.1. We'll see how that goes... Sorry to hear that. But I thought of one more thing to check after you get everything running again. Double click on the Network icon on the desktop. (If that isn't there, open up Windows Explorer and click on the Network icon in the left window.) See what devices and computers show up. A screen capture might be useful. You should definitely see the unRAID server and your Windows 7 computer... Then click on you server and see if you can see all of your shares and your flash drive. A list of your server hardware could be useful. I looked through your diagnostics file and I saw a lot of "ATA bus error" messages in there. And lines like this: Oct 15 18:27:16 Tower shfs/user: err: shfs_rmdir: rmdir: /mnt/cache/SABnzbd/incomplete/Terri.2011.BluRay.1080p.AVC.DTS-HD.MA5.1-CHDBits.cp(tt1687281) (39) Directory not empty I am no expert on syslogs but something does not look right...
October 18, 20169 yr Author That conf file did force Tower to be Master, but it didn't solve the problem. Still getting the same failed file transfers to and from the Tower On the Windows computer, go to 'Control Panel', then 'Credential Manager' and see if you have any credentials set. (However, if 6.1 was working, I find it hard to believe that this is the issue but it is quick and easy to check.) Do you have any other computers that you could use to help isolate the problem? Have you changed the 'SMB Security Settings' from the defaults? The reason I am puzzled is that most of these types of issues have been occurring with the Windows 10 OS. Windows 7 has been pretty much trouble free... To add insult to injury, while testing back and forth between builds, the USB drive died. All USB fixer tools have reported it unrepairable. Now I need to wait a day for my new USB and then I'll try a fresh built of 6.2.1. We'll see how that goes... Sorry to hear that. But I thought of one more thing to check after you get everything running again. Double click on the Network icon on the desktop. (If that isn't there, open up Windows Explorer and click on the Network icon in the left window.) See what devices and computers show up. A screen capture might be useful. You should definitely see the unRAID server and your Windows 7 computer... Then click on you server and see if you can see all of your shares and your flash drive. A list of your server hardware could be useful. I looked through your diagnostics file and I saw a lot of "ATA bus error" messages in there. And lines like this: Oct 15 18:27:16 Tower shfs/user: err: shfs_rmdir: rmdir: /mnt/cache/SABnzbd/incomplete/Terri.2011.BluRay.1080p.AVC.DTS-HD.MA5.1-CHDBits.cp(tt1687281) (39) Directory not empty I am no expert on syslogs but something does not look right... I'll look into the ATA bus errors. I think the rmdir errors were from trying to remove appdata folder stuff that has write permission restrictions built into the functionality
October 18, 20169 yr Author After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server?
October 18, 20169 yr Author After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server? Durrrrr. Problem is still happening. Just occurred on simultaneous writes to the server.
October 18, 20169 yr After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server? Durrrrr. Problem is still happening. Just occurred on simultaneous writes to the server. Please, when you post about a problem, no one can offer help unless you explain what you did that made the problem happen, and if possible, capture the diagnostics of the server state while the problem is happening or right after (that is, without server reboot). In your first log, there are ATA errors on one of your drives. This is one issue to address. Also saw apcupsd fire onto battery and off again. What do you know about this? Maybe you have flaky power?
October 18, 20169 yr Author After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server? Durrrrr. Problem is still happening. Just occurred on simultaneous writes to the server. Please, when you post about a problem, no one can offer help unless you explain what you did that made the problem happen, and if possible, capture the diagnostics of the server state while the problem is happening or right after (that is, without server reboot). In your first log, there are ATA errors on one of your drives. This is one issue to address. Also saw apcupsd fire onto battery and off again. What do you know about this? Maybe you have flaky power? This thread already contains full diagnostics, however, I've attached fresh diagnostics to this post. If I have a power or ATA issues, I'll deal with them later, in a separate thread -- the PSU is brand new and more than sufficient for the load. The read/write problems occur only after after updating. If I downgrade to 6.1.9, the problem goes away. Not sure how to address... The problem is simple, if I try to do more than one simultaneous file transfer to/from the server to my local machines, the transfers fail with errors about the Specified Network Name is No Longer Available. tower-diagnostics-20161018-1819.zip
October 18, 20169 yr Community Expert I remember having this exact issue a few years back, not with unRAID but when copying files from windows 7 to windows 2012 server, after googling and finding other users with same issue but no solution that worked for me I "solved it" by upgrading my desktop to windows 8.1, although I was not a big fan at the time, but the problem was indeed solved. This may not be a solution for you but I suspect if you have other desktop/laptop running windows 8.1/10 you won't see this error, IIRC this issue was more prevalent between windows 7 (SMB2) and SMB3 computers, unRAID now also uses SMB3.
October 18, 20169 yr Author I remember having this exact issue a few years back, not with unRAID but when copying files from windows 7 to windows 2012 server, after googling and finding other users with same issue but no solution that worked for me I "solved it" by upgrading my desktop to windows 8.1, although I was not a big fan at the time, but the problem was indeed solved. This may not be a solution for you but I suspect if you have other desktop/laptop running windows 8.1/10 you won't see this error, IIRC this issue was more prevalent between windows 7 (SMB2) and SMB3 computers, unRAID now also uses SMB3. Thanks. Sadly, I'm on 100% Windows 7 machines with no interest in upgrading.
October 18, 20169 yr After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server? Durrrrr. Problem is still happening. Just occurred on simultaneous writes to the server. Please, when you post about a problem, no one can offer help unless you explain what you did that made the problem happen, and if possible, capture the diagnostics of the server state while the problem is happening or right after (that is, without server reboot). In your first log, there are ATA errors on one of your drives. This is one issue to address. Also saw apcupsd fire onto battery and off again. What do you know about this? Maybe you have flaky power? This thread already contains full diagnostics, however, I've attached fresh diagnostics to this post. If I have a power or ATA issues, I'll deal with them later, in a separate thread -- the PSU is brand new and more than sufficient for the load. The read/write problems occur only after after updating. If I downgrade to 6.1.9, the problem goes away. Not sure how to address... The problem is simple, if I try to do more than one simultaneous file transfer to/from the server to my local machines, the transfers fail with errors about the Specified Network Name is No Longer Available. I tried 3 simultaneous large file transfer from a win7 desktop to 3 different shares on a server: disk share, cache share, user share - no problems. This is just using windows explorer to drag the files over. There is definitely a problem with this device: Oct 18 16:54:37 Tower kernel: ata1.00: ATA-9: Samsung SSD 850 PRO 1TB, S2BBNWAG118877N, EXM02B6Q, max UDMA/133 This the syslog message: Oct 18 18:08:12 Tower kernel: ata1.00: exception Emask 0x10 SAct 0x3000 SErr 0x280100 action 0x6 frozen Oct 18 18:08:12 Tower kernel: ata1.00: irq_stat 0x08000000, interface fatal error Oct 18 18:08:12 Tower kernel: ata1: SError: { UnrecovData 10B8B BadCRC } Oct 18 18:08:12 Tower kernel: ata1.00: failed command: READ FPDMA QUEUED Oct 18 18:08:12 Tower kernel: ata1.00: cmd 60/00:60:c0:29:af/08:00:01:00:00/40 tag 12 ncq 1048576 in Oct 18 18:08:12 Tower kernel: res 40/00:68:c0:31:af/00:00:01:00:00/40 Emask 0x10 (ATA bus error) Oct 18 18:08:12 Tower kernel: ata1.00: status: { DRDY } Oct 18 18:08:12 Tower kernel: ata1.00: failed command: READ FPDMA QUEUED Oct 18 18:08:12 Tower kernel: ata1.00: cmd 60/00:68:c0:31:af/08:00:01:00:00/40 tag 13 ncq 1048576 in Oct 18 18:08:12 Tower kernel: res 40/00:68:c0:31:af/00:00:01:00:00/40 Emask 0x10 (ATA bus error) Oct 18 18:08:12 Tower kernel: ata1.00: status: { DRDY } Oct 18 18:08:12 Tower kernel: ata1: hard resetting link DMA errors with BadCRC for hard drives almost always means bad cable, bad controller or bad power. Might be different with that SSD. I don't know why it works with 6.1.9 and not 6.2. Lots of s/w changed, different memory is being used. You have a new PSU - it's possible it's marginal. Did you run a pass of memtest (though memtest does not always find bad memory)? Note: For those reading: why would memtest not find all bad memory? Because it only tests the data path between the RAM chips and the CPU. It does not test the data path from DMA controller (PCI bus) to RAM chips. If there is a bad memory cell itself, it will find it, but if there is a problem in some other circuit it might not.
October 18, 20169 yr Author After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server? Durrrrr. Problem is still happening. Just occurred on simultaneous writes to the server. Please, when you post about a problem, no one can offer help unless you explain what you did that made the problem happen, and if possible, capture the diagnostics of the server state while the problem is happening or right after (that is, without server reboot). In your first log, there are ATA errors on one of your drives. This is one issue to address. Also saw apcupsd fire onto battery and off again. What do you know about this? Maybe you have flaky power? This thread already contains full diagnostics, however, I've attached fresh diagnostics to this post. If I have a power or ATA issues, I'll deal with them later, in a separate thread -- the PSU is brand new and more than sufficient for the load. The read/write problems occur only after after updating. If I downgrade to 6.1.9, the problem goes away. Not sure how to address... The problem is simple, if I try to do more than one simultaneous file transfer to/from the server to my local machines, the transfers fail with errors about the Specified Network Name is No Longer Available. I tried 3 simultaneous large file transfer from a win7 desktop to 3 different shares on a server: disk share, cache share, user share - no problems. This is just using windows explorer to drag the files over. There is definitely a problem with this device: Oct 18 16:54:37 Tower kernel: ata1.00: ATA-9: Samsung SSD 850 PRO 1TB, S2BBNWAG118877N, EXM02B6Q, max UDMA/133 This the syslog message: Oct 18 18:08:12 Tower kernel: ata1.00: exception Emask 0x10 SAct 0x3000 SErr 0x280100 action 0x6 frozen Oct 18 18:08:12 Tower kernel: ata1.00: irq_stat 0x08000000, interface fatal error Oct 18 18:08:12 Tower kernel: ata1: SError: { UnrecovData 10B8B BadCRC } Oct 18 18:08:12 Tower kernel: ata1.00: failed command: READ FPDMA QUEUED Oct 18 18:08:12 Tower kernel: ata1.00: cmd 60/00:60:c0:29:af/08:00:01:00:00/40 tag 12 ncq 1048576 in Oct 18 18:08:12 Tower kernel: res 40/00:68:c0:31:af/00:00:01:00:00/40 Emask 0x10 (ATA bus error) Oct 18 18:08:12 Tower kernel: ata1.00: status: { DRDY } Oct 18 18:08:12 Tower kernel: ata1.00: failed command: READ FPDMA QUEUED Oct 18 18:08:12 Tower kernel: ata1.00: cmd 60/00:68:c0:31:af/08:00:01:00:00/40 tag 13 ncq 1048576 in Oct 18 18:08:12 Tower kernel: res 40/00:68:c0:31:af/00:00:01:00:00/40 Emask 0x10 (ATA bus error) Oct 18 18:08:12 Tower kernel: ata1.00: status: { DRDY } Oct 18 18:08:12 Tower kernel: ata1: hard resetting link DMA errors with BadCRC for hard drives almost always means bad cable, bad controller or bad power. Might be different with that SSD. I don't know why it works with 6.1.9 and not 6.2. Lots of s/w changed, different memory is being used. You have a new PSU - it's possible it's marginal. Did you run a pass of memtest (though memtest does not always find bad memory)? Note: For those reading: why would memtest not find all bad memory? Because it only tests the data path between the RAM chips and the CPU. It does not test the data path from DMA controller (PCI bus) to RAM chips. If there is a bad memory cell itself, it will find it, but if there is a problem in some other circuit it might not. Thanks. Are you saying the SSD is bad/should be replaced as a starting point? Sent from my HTC6535LVW using Tapatalk
October 18, 20169 yr Author After replacing the USB thumb drive, I haven't been able to recreate the problem. Any logical reason that a failing flash drive would have created read/write errors to and from the server? Durrrrr. Problem is still happening. Just occurred on simultaneous writes to the server. Please, when you post about a problem, no one can offer help unless you explain what you did that made the problem happen, and if possible, capture the diagnostics of the server state while the problem is happening or right after (that is, without server reboot). In your first log, there are ATA errors on one of your drives. This is one issue to address. Also saw apcupsd fire onto battery and off again. What do you know about this? Maybe you have flaky power? This thread already contains full diagnostics, however, I've attached fresh diagnostics to this post. If I have a power or ATA issues, I'll deal with them later, in a separate thread -- the PSU is brand new and more than sufficient for the load. The read/write problems occur only after after updating. If I downgrade to 6.1.9, the problem goes away. Not sure how to address... The problem is simple, if I try to do more than one simultaneous file transfer to/from the server to my local machines, the transfers fail with errors about the Specified Network Name is No Longer Available. I tried 3 simultaneous large file transfer from a win7 desktop to 3 different shares on a server: disk share, cache share, user share - no problems. This is just using windows explorer to drag the files over. There is definitely a problem with this device: Oct 18 16:54:37 Tower kernel: ata1.00: ATA-9: Samsung SSD 850 PRO 1TB, S2BBNWAG118877N, EXM02B6Q, max UDMA/133 This the syslog message: Oct 18 18:08:12 Tower kernel: ata1.00: exception Emask 0x10 SAct 0x3000 SErr 0x280100 action 0x6 frozen Oct 18 18:08:12 Tower kernel: ata1.00: irq_stat 0x08000000, interface fatal error Oct 18 18:08:12 Tower kernel: ata1: SError: { UnrecovData 10B8B BadCRC } Oct 18 18:08:12 Tower kernel: ata1.00: failed command: READ FPDMA QUEUED Oct 18 18:08:12 Tower kernel: ata1.00: cmd 60/00:60:c0:29:af/08:00:01:00:00/40 tag 12 ncq 1048576 in Oct 18 18:08:12 Tower kernel: res 40/00:68:c0:31:af/00:00:01:00:00/40 Emask 0x10 (ATA bus error) Oct 18 18:08:12 Tower kernel: ata1.00: status: { DRDY } Oct 18 18:08:12 Tower kernel: ata1.00: failed command: READ FPDMA QUEUED Oct 18 18:08:12 Tower kernel: ata1.00: cmd 60/00:68:c0:31:af/08:00:01:00:00/40 tag 13 ncq 1048576 in Oct 18 18:08:12 Tower kernel: res 40/00:68:c0:31:af/00:00:01:00:00/40 Emask 0x10 (ATA bus error) Oct 18 18:08:12 Tower kernel: ata1.00: status: { DRDY } Oct 18 18:08:12 Tower kernel: ata1: hard resetting link DMA errors with BadCRC for hard drives almost always means bad cable, bad controller or bad power. Might be different with that SSD. I don't know why it works with 6.1.9 and not 6.2. Lots of s/w changed, different memory is being used. You have a new PSU - it's possible it's marginal. Did you run a pass of memtest (though memtest does not always find bad memory)? Note: For those reading: why would memtest not find all bad memory? Because it only tests the data path between the RAM chips and the CPU. It does not test the data path from DMA controller (PCI bus) to RAM chips. If there is a bad memory cell itself, it will find it, but if there is a problem in some other circuit it might not. Thanks. Are you saying the SSD is bad/should be replaced as a starting point? Sent from my HTC6535LVW using Tapatalk That's in a breakout slot for a 21st drive. I can try it in a standard drive cage, too... Or I could try disabling writes to the Cache drive, and see if multiple writes at once are successful? Sent from my HTC6535LVW using Tapatalk
Archived
This topic is now archived and is closed to further replies.